Research Questions
January 4, 2024
Questions have a alluring quality to them. Each one is like an invitation. A call to adventure. As an end of year reflection, I pulled out every question (lightly curated) that I wrote in my research log this past year. Reading through them reminded me of the emotional and intellectual forces that drove my work this year–many still alive in me now. I’m sharing them here in case they spark something for you too ✨
First impression, what was interesting to you and why did you add it?
What questions might be answered by reading it that are relevant to our work?
Can you help someone get from where they are to new territory?
Interaction design is as much about connecting humans across the murky “Internet cloud” (fostering community and conversation) as connecting an individual with his or her own capacity to explore what is possible and generate new possibilities (supporting internal conversations).” Hugh Dubberly in [What is Conversation?
Once you’ve narrowed to one persona, we can get more specific about their scenarios and try to identify more clearly: what does a theme mean to this person?
What kinds of operations do they need to perform on a theme?
How do those operations relate to their broader context?
One thing that seems to be missing from all these scenarios is: “then what?” How does the activity you’re describing (manipulating notes and themes) actually connect to the broader goals and behaviors you articulated?
For instance, what (if any) behavioral/cognitive theory will underpin our design ideas given this new problem/scenario?
Klein’s sensemaking framework is still partially relevant, but not in the same ways… how does it fit in?
Can we articulate any other theoretical frameworks (possibly our own) that will inform our design besides intuition and user testing?
How does that feel to you?
If we design primarily for this persona, will we be able to appropriately evaluate design ideas ourselves?
Maybe you’re thinking that the analogue in our context is us getting each other unstuck in this project?
Does that roughly match what you’re thinking?
Of course helping with communication will be great, too, but it seems like step two?
Would changing their supplier allow them to get around some of the tradeoffs?
How does that fit into the scenario you imagine?
Are there other “multiplayer” elements that interact with the system here, apart from using the system to construct some kind of distilled/structured summary to send others?
In what ways?
What has been mentioned/explored/discussed so far?
What is the big picture?
Missing?
What problem is this solving exactly?
What is the nature of the problem?
How can all the chaff be used constructively to make the new thing?
What is hard about using all the working material as fodder for progress?
Is there a suite of mechanisms like this, or is this the best one?
How does this mechanism relate to the problem statement?
What are the facets of the paragraph being emphasized?
Can semantic search be valuable for highly unstructured text corpora?
What is the mechanism that’s being shown here?
What is the reason for using this conceptual design?
What problem is the conceptual design solving?
How can semantic search be improved for better generation results?
What is the value of semantic search for unstructured text corpora?
How can the mechanism be tied to the experience in Japan?
What is the problem that the conceptual design is designed to solve?
What are the implications of the conceptual design for changing the problem?
How long is the current runway for the project?
What opportunities may arise as a result of the project?
What will happen in the next year based on the project’s progress?
What are some potential future directions for the project?
What is the essence of the problem statement being addressed?
How can the mechanism be made more userfriendly?
One way to test a problem statement is to step back and ask: does this seem to actually match the sorts of design ideas you find yourself interested in generating?
And: would this problem, if solved, be the thing your persona really needs?
What is that something?
What does he need to be able to do to solve his problem with those materials?
“Review and remix”?
But it seems to me that this problem statement would also authorize a completely different direction focused on, say, semantic search—wouldn’t that help make the material “accessible” for “review”?
How should we decide whether these two ideas or the themedistillation idea is the right path?
What is that something?
What does he need to be able to do to solve his problem with those materials?
Reconnect with the emotional and intellectual force that drove the production of those materials?
Are they still salient now?
What was most salient to him at that time and past context?
What is still salient from that past context?
What will form will recreate the most vivid mental image of a past activity?
what do you need to do when looking at this material?
materials produced/gathered that may be relevant on multiple time scales, losing emotional connection to work; What were the things that invigorated the work initially?
How does it feel to use this?
How similar do you want the sequence of prompts to be?
Love, wisdom, grace, inspiration — how do you go about finding these things that are in some ways about extending the boundaries of the self into unknown territory, about becoming someone else?
What form will recreate the most vivid mental image of a past activity?
When you’re in a situation, what will you use?
How malleable are these dispositions?
How to push this further this week?
How to balance push and pull?
How do we show learning?
Having progress indicators that counterbalance each other?
Quote 2: “So then, I guess the question that I need to explore a little bit more with Andy this week is like wha is the purpose of memory?
How would tove jansson (propn) use (v) spaced repetition (n)?
Maybe it would be interesting to have a little conversation about a prompt, as an exercise?
What are some powerful visions of computing from the past?
What did they do well and where did they struggle?
how do you enable talent to do the things they want to do?
Asked, what would happen if in ten years you were told to stop using the device?
A reflection/memory/creativity support system that in ten years, would be something people would be unwilling to put down?
What would it look like?
What does the space of information visualizations look like?
How can we develop a meaningful human augmentation laboratory?
How can we develop transformative tools for thought?
How did the computer animation industry get started; was it public research?
Seems like Pixar was able to build on their work?
Perhaps progress can measured by the rate which new fields are created?
What role does emotion play in memory and cognition?
If spaced repetition were different genres of video game, how would you play?
What does it take to become an expert in Chess?
What are the differences between a novice and an expert in a field?
What happened in that minute of time?
What things did they pay attention to, what cues and signals helped them arrive at that answer?
And is it possible to externalize and teach those cues to novices?
Programmable attention?
Expanding the variety of prompts that people can get in a spaced repetition system?
What happens when people design or are prompted with questions that come from these categories?
Helping people navigate the messy middle of a creative project by remixing materials (concepts?
So is this is this is this process useful?
How do people do this process today?
tools for thought how do we amplify and support human intelligence and understanding?
concepts are really powerful, like, what other strategies does Taiwan have access to?
Why does a chair look like this?
Why do we have these particular tools?
Why is a T shirt printed with the things that it has on it?
What is it communicating?
What is it saying about this world?
And then you’re asking yourself, How do I feel about this?
Like, what does it does it does it adequately solve or like communicate the thing that I’m trying to say?
Or is there a clear way to say the thing I want to say that doesn’t exist already?
We need experience to validate the components of a concept how do you know what a profit generating home looks like?
Does anyone else experienced this?
And are we together in this industry experience that we call life which?
What are some of the hard parts of developing a concept?
What comes to mind?
And so how do we support the human in this process?
And so how do we support the human solving problems?
How do we support the human working on complex situations?
What are the hard parts about creating concepts?
What does it take to develop a concept or an abstraction?
Why?
Why?
Why?
And so, how do we discover these patterns?
And we need to enrich it first we need to win first we need to deconstruct it we need to understand the pieces of reality like when this conversation with as I’m speaking what are the things that I said what are the raw things are the things that are important?
The RE mixing problem and helping us juggle concepts up today, and yeah, I guess this is an open question of what is the unit of knowledge that is important to work with?
What do I want to do with references?
How have people done online references in the past?
Is there some way to evaluate the quality of an abstraction/waypoint?
for example, the amount of hierarchical nesting?
How have people designed pivot systems in the past?
Can we discover and explain highlevel patterns?
Do they generate insights that guide our design?
Do they spark ideas?
Do they lead to great systems?
Does it make sense to have all these sorts of views that you can select from?
What is the commonality between these things?
But if we are able to choose where this comes from, from a place we have deemed meaningful, how lucky are we?
Can you understand different cultures by how their understanding of various concepts differ?
Why do we need concepts and how do we use them?
Similar how?
When do we need concepts?
Why do I think figuring out what to do next can be described as a concept problem?
Why is this concept important to me?
What do I know about this concept?
What do I want to know about this concept?
Why did I start thinking about this concepts stuff?
What was the point?
What is the current problem statement?
Why is this concept important to me?
What do I know about this concept?
What do I want to know about this concept?
How does this affect what I’m doing?
Why is a concept important to me?
At some point, you’re like why did I start reading in the first place?
Did I finish it he gathering or do I still have more to do?
Does research confirm this?
How long should students persist to find the answers?
Is this time efficiently spent?
Or, would it be better to have the students try to answer a few questions, pool their information as a class, and then move to practicing retrieval of the information?
What are the important concepts from this conversation?
Ex: Why do we need concepts?
What is a context scenario?
How does the proposed system map to Klein and sensemaking?
What are the differences between Andy and my own conception of what it means to figure out what to do next?
What information do we keep and why do we keep it?
For example, “that’s interesting”, “tell me more”, “how did that make you feel”?
Do users want to make copies of highlights and out them in multiple places?
For example, “that’s interesting”, “tell me more”, “how did that make you feel”?
What set functions are most important?
For example, are they typically native or nonnative?
What about the type of bark?
What method do they use to reproduce?
Are they seed or fruit bearing?
How we can mix ideas together like chemical elements to produce new ones?
How can we make the computational medium more malleable?
What if we could deconstruct these pages into its components like Lego blocks?
Then remix and reuse these components to construct new views to answer new questions?
How do we navigate treelike structures?
When all mediums of communication can be represented by dense vectors, what does this enable?
What does the zoomed out view for text look like?
What if we rendered the web like a game engine?
Can we use brain imaging technologies to understand how various cognitive exercises activate the cortex?
What are the strengths and weaknesses of spoken and written language?
Radical independence what if concepts were truly independent?
What would that look like?
If you don’t care as much about moving stuff around the page, what could you do?
What is a block?
What ways can you reference text in other mediums?
What if there was something similar with interfaces, systematically reintroducing elements back?
Where can I learn more?
How do I want to convey my interests, work and past experiences?
My sense is that we’ve gestured at some possiblygeneral powerful ideas “explainable” LLMbased summarization can augment frame formation?
Naively, I wonder: is there some version of this idea that you would be more familiar with?
Or, relatedly, why are you personally so excited about these directions in the first place?
Isn’t it because (even before this project) you’ve found yourself in situations where something like it would be needed?
Can you draw on those situations?
Relatedly, I also wonder if extracting signal from interview or conversation transcripts might feel more familiar to you than your logbook?
Then at the moment that you’ve identified a salient reframe, how quickly can you assemble that frame from the existing material?
why are you personally so excited about these directions in the first place?
Framing these beliefs as a question, how can we develop more reflective mediums for wayfinding?
why are you personally so excited about these directions in the first place?
Posed as a question, how can we develop powerful, new mediums for wayfinding?
how can we develop powerful, new mediums for wayfinding?
I find it funny that I can write something with such a high level of conviction, only to look at it later and ask, was that really what mattered?
how can we develop powerful, new mediums for wayfinding?
Posed as a question, how can we develop powerful, new mediums for wayfinding?
Is there a connection to life as improv?
What problem am I solving for?
What’s so great about the digital medium anyway?
to what extent is achieving that goal about learning to write better prompts?
Are individual prompts the best primitives for that task?
If we step back and look at what’s actually necessary to produce that sort of understanding, and at the mechanisms of practice, could we find a bettermatched primitive?
Reframing “memory systems” as “practice systems” for a moment, is there some other kind of primitive waiting to be created—one which would let me “add the concept” as an elementary operation, in the same way I can “add a vocabulary word” to one of today’s memory systems?
What does it mean to “add a concept”?
What specifically would you be adding?
What is research?
What other verbs could be useful for nonmemory prompts like these?
However, perhaps it’s worth another look?
What is the problem with working with these brain dumps?
Grab all the salient items, put them in one place, then ask what makes these things related?
what are they trying to tell me?
Ok so working memory, you’re feeling overwhelmed, why?
What was wrong about what he did here?
Are we saying could have done this faster?
Or easier?
Or better?
Layer: Interface?
And then if someone asks how come?
What is the challenge with making these handles?
How do we make those handles today?
Why do we need handles?
For example I have a section called what is the problem with working with these brain dumps?
What do concepts do for you?
Which ones have the most use across contexts?
Why was this relevant?
Why do we develop them?
An open question, is how do abstract concepts develop?
What are the challenges with developing concepts?
Is it getting them socially agreed upon?
What words are real and what aren’t?
Are these words real, really asking how many brains will this give me access to?
Is language the ultimate artefact?
What exactly are the unaided cognitive capacities of the human brain?
We have this dense mess of text and we want to ask ourselves what was important?
How do we know what to care about?
But what does it say?
What is the challenge with making these handles?
Why do we create abstractions?
How do we develop abstractions?
What is important to consider?
What was the one thing that shifted today?
So where have I been getting stuck?
What was the problem that it solves for?
Why do people take the action?
What is the artifact we’re trying to create?
Can we discover and explain highlevel patterns?
Do they generate insights that guide our design?
Do they spark ideas?
Do they lead to great systems?
Why?
How?
What do I know about pivots?
What makes them so useful?
What do I want to know more about them?
Why is this pattern powerful?
What implications does this have on my goal if it’s true?
So is the goal to invent new definitions?
New words to put in some local dictionary?
Or inventing new definitions for words to be used in some specific context that emphasizes different things?
Or new phrases?
What do they share in common?
For example, ok what seems emotionally important from what I’m saying today?
Ok, why do I feel like that?
Ok great, what do I know about pivots?
What would help me implement these better?
Why might pivots not work?
Now why is it interesting?
Like are they going somewhere?
Importance scale why do I care about this concept?
Why do I feel this concept is important?
Why might this concept not be important?
What do I know about this concept?
What do I want to know about this concept?
What are you pivoting on?
Why might this concept not be important?
What other moves should I make?
Bret Victor’s thoughts on abstraction levels?
How to process this?
Who else has thoughts on this?
Why do we create abstractions?
What type of abstractions do I find myself creating?
What is difficult about creating abstractions?
What is the problem with working with these brain dumps?
Grab all the salient items, put them in one place, then ask what makes these things related?
what are they trying to tell me?
Layer: Interface?
I think you can probably help yourself by sharpening up “dealing with large quantities of nove information.” What does “dealing with” mean here?
Are they generating the information?
Collecting it from other people?
How does this first sentence connect to the personal, reflective, emotional emphasis you’ve been chasing for the past couple weeks?
Is it too broad to say both generating and collecting?
Are we focused on researchers here?
And for intellectual purposes, like you’re describing, rather than personal emotional ones?
I’d suggest we maybe spend an hour tightening it up in realtime?
Would you be game to meet after the call with Dan tomorrow?
Is it too broad to say both generating and collecting in regards to researcher dealing with information?
is there anything specific I do that is particularly effective or you’ve seen elsewhere?
What cognitive tasks are various actions solving?
Why are you making the moves that you are doing?
Wasn’t that what we were doing at the start?
Going by gut intuition?
GPT: What is my primary goal or purpose in engaging with these theories, practices, and interviews?
What do I ultimately want to achieve?
Could we just move forward and develop this technique for ourselves?
Then see where other people choose to use it?
What type of paper do I imagine?
Not sure how I can do this if it goes against his beliefs in LLMs?
But are there potentially more powerful memory techniques that we can leverage?
Ones that people might not use as much today because they are time intensive and difficult to do?
Given the computer’s ability to process information, can we use it to help us create these powerful artifacts?
Working memory and cognitive load are insufficient, for what?
Why don’t current processes work?
Why do you need to notice clusters?
Is it too broad to say both generating and collecting in regards to researcher dealing with information?
Human computer symbiosis?
What would a game about exploring your mind look like?
What is the goal of such a game?
So what did you do today?
How did that connect to your goals?
Why?
Miyamoto asking, what is the logical way to kill enemies in a game about jumping?
One of the main ones being, what if I don’t find anything powerful here?
What were the powerful phrases I was playing with?
What other things would I do if not this?
But for what?
Maybe I could work on something else?
Then what’s really the difference this time?
what if you couldn’t access your past writings?
What if the only way to do so was to unlock a portal?
What other things would I do if not this?
What’s really the difference this time in working on the project?
Consider the possibility of working on something else or with different people: “Maybe I could work on something else?
Question: How about the Default mode network?
Does slower thinking, like handwriting vs. typing lead to more creative insights?
What is the word for seeing your reflection in the water?
What if everyone had the space they needed to become?
What is the optimal level of structure in a journaling practice for maximum personal growth and introspection?
How can technology and tools like text analysis, sentiment analysis, and machine learning algorithms be used to enhance introspective journaling and review of past entries?
This could be something like what are the risks?
So does it mean it didn’t feel important?
What makes an introspection practice powerful and what does it usually take to get there?
Why does the artist’s way work so well?
User extracted main points?
Interface supported extraction?
Do they affect working memory capacity?
What are the distinct practices?
How does each practice vary?
What are their learning objectives?
What are the distinct challenges in each?
How do they figure out what they think after ranting for an hour?
What is actually happening?
What am I to do today?
Harder to deal with the artifact of the audio?
are there any particular moves?
How do you know when you’re done?
What would be the stupider way to do this?
How can journaling best be incorporated into daily routines to maximize consistency and reflection?
What are the most effective techniques for extracting insights and patterns from journal entries?
How can individuals find the most effective journaling practices for their specific needs and interests?
What strategies can be employed to maintain a consistent journaling routine over time?
What is the hard thing about making sense of brain dumps or stream of consciousness writing?
Like if we created some intervention, designed some interface, enabled this type of action to be taken, how would we know that things have improved?
What are we comparing that intervention against?
Have you even become aware of the right action?
Is it in the set of actions that you are considering?
How do I even start to think about this problem?
How often do you encounter recurring themes, patterns or issues while writing or reviewing your work?
Can you share an example of a time when you struggled to identify or analyze patterns in your writing?
What was the challenge?
How much time do you typically spend summarizing, gathering supporting evidence, and categorizing your thoughts while writing?
Have you ever used any tools or techniques to help you identify patterns in your writing?
If so, what were they and how effective were they?
In your opinion, how important is it to be able to effectively identify patterns in your writing?
Why?
How do you believe a computerbased solution could assist you in identifying patterns and gathering supporting evidence in your writing?
Are there any specific features or capabilities you would like to see in a tool that helps you identify patterns in your writing?
How much time do you think you could save if you had a tool that could automatically suggest categorizations and gather supporting evidence for you?
Do you think that having a tool that can identify patterns and categorizations in your writing would improve your overall writing quality?
Why or why not?
Are there any concerns or potential drawbacks you foresee with using a computer based solution to identify patterns and categorizations in your writing?
Can you briefly describe your experience with writing supported thinking or problem solving sessions?
Have you ever struggled with setting clear objectives or goals at the beginning of a writing supported thinking session?
Can you provide an example?
How important do you think it is to have clear objectives when engaging in a writing supported thinking session?
Why?
Have you ever faced challenges in identifying the right goals or questions to focus on during a writing supported thinking session?
How did you overcome this challenge?
Do you have any strategies or techniques for setting clear objectives prior to or during a writing supported thinking session?
How do you think setting clear objectives early in a writing supported thinking session might impact your ability to generate actionable insights or solutions?
Can you recall a situation where you were unable to set clear objectives in a writing supported thinking session, and it affected the outcome?
What happened, and what did you learn from the experience?
In your opinion, what are the key components of effective objectives for a writing supported thinking session?
Do you have any recommendations for improving objectivesetting in writing supported thinking sessions?
How often do you engage in writing or brainstorming sessions to organize your thoughts and ideas?
Have you ever experienced difficulty in identifying and prioritizing the most important information from your brainstorming sessions or extensive notes?
How much time do you usually spend on curating and organizing your notes or ideas after a brainstorming session?
Are there any specific methods or tools you currently use to help you curate and prioritize information from your notes or brain dumps?
Have you ever felt that you missed out on important information or ideas because of the timeconsuming nature of curating and organizing your notes?
Would you be interested in a tool that can help you annotate, curate, and prioritize information based on your preferences and examples?
Can you share any instances where you felt overwhelmed or struggled to make sense of your notes and wished for a more efficient way to curate the content?
How important is it for you to have a wellorganized and prioritized set of notes or ideas for your personal or professional life?
Are there any specific features you would like to see in a tool or system that helps you curate and prioritize information from your notes?
Do you believe that a computeraided system could effectively help you in annotating and curating your notes, or do you think it’s a task best done manually?
Can you describe a situation when you felt that your individual or group perspective might have been limited due to cognitive biases?
Have you ever experienced a situation where incorporating external feedback helped you overcome a cognitive bias?
Can you think of any examples where your team struggled to consider all possible solutions to a problem?
What do you think caused this limitation in your thinking?
When faced with a complex problem, how do you ensure that you’re considering the largest possible subset of solutions?
How do you identify potential biases in your own thinking, or within your team’s thought processes?
Have you ever used prompts or tools like Oblique Strategies or Bloom’s Taxonomy to help explore different perspectives on a problem?
If so, can you provide an example and describe the results?
How open are you and your team to receiving external feedback or suggestions?
What methods do you use to seek out such feedback?
Do you believe that incorporating external feedback or utilizing tools like Oblique Strategies can help individuals and teams achieve better coverage of the possibility space?
Why or why not?
How do you handle situations where there might be a lack of available external feedback?
What strategies do you use to ensure you’re still considering a wide range of possibilities?
In your opinion, what are the most common cognitive biases that limit an individual’s or a group’s ability to fully explore the possibility space?
How can we overcome these biases?
Do you regularly engage in writing or journaling as a means of selfreflection and personal growth?
How often do you revisit your past writings or journal entries?
Have you ever found it challenging to remember specific prompts or thoughts from previous writing sessions that you would like to revisit?
Can you provide an example of a time when revisiting a past writing prompt or thought would have been beneficial for your current situation?
How do you currently keep track of important prompts, thoughts, or ideas from your writing sessions that you may want to revisit in the future?
Would you find value in a tool that periodically reminds you of salient information from your past writings, prompts, or thoughts?
If such a tool existed, how often would you want to be reminded of past prompts or thoughts?
In what situations or contexts would you find it most useful to be reminded of past prompts or thoughts?
Do you have any concerns or potential issues with using a computerbased tool to help you remember and revisit past writing prompts or thoughts?
Are there any additional features or capabilities you would like to see in a tool designed to help you periodically revisit salient information from your past writings?
Can you tell me about a time where you felt the journaling practice was really good?
How about not good?
How do your journaling sessions begin?
What prompts them?
How do you know you’re done?
Is there anything you keep track of over time?
Do you revisit past writings or journal entries?
What would you want to revisit?
Do you journal to work through problems in your life?
What do you do if you get stuck in your journal?
How often do you engage in writing or brainstorming sessions to organize your thoughts and ideas?
Can you share any instances where you felt overwhelmed or struggled to make sense of your notes and wished for a more efficient way to curate the content?
How can a computerbased solution assist in identifying patterns and gathering supporting evidence in writing?
What are the key components of effective objectives for a writingsupported thinking session?
Are there any specific features desired in a tool or system that helps curate and prioritize information from notes?
How can we overcome common cognitive biases that limit the exploration of the possibility space?
What are the most useful situations or contexts for being reminded of past prompts or thoughts?
systems?
For example, how might shared representations help promote interpretability, learning, and skill acquisition by more novice users, using AI methods to level up—rather than deskill— workers?
To what degree will such systems promote behavior characterized by user control vs. more passive acceptance of algorithmic recommendations?
Can appropriate designs prompt more critical engagement in the face of automated support?
In some cases, should we accept having the computer “think for us,” if it frees us to engage in higherlevel reasoning or more creative practice?
Once we have these directions, there’s a very natural question to ask: which words are furthest in these directions?
What are the most masculine or feminine words?
The most adult, the most childish?
To what degree will such systems promote behavior characterized by user control vs. more passive acceptance of algorithmic recommendations?
Can appropriate designs prompt more critical engagement in the face of automated support?
In some cases, should we accept having the computer “think for us,” if it frees us to engage in higherlevel reasoning or more creative practice?
What prompts people to write?
Am I doing the right things?
What should I do next?
What makes me feel alive?
What do I want to get done today?
What scares me right now?
What makes me feel alive?
What am I grateful for?
What am I proud of today?
How am I spending my days?
Like in a conversation, someone telling you about how remember you were so scared of doing these types of things before?
What are the feelings I want people to experience?
So where do stories come from?
So can a technology do this for ourselves?
What are the feelings we want to feel?
Why was that idea so popular?
Salience over what?
Do I seem angry?
angry over what?
What do you think?
What types of patterns can you extract from key presses?
Has anyone done this?
So can personal analytics help me measure the rate at which that happens?
What motivated you to start a reflective practice like lifelogging or journaling?
How long have you been engaging in your reflective practice, and how has it evolved over time?
What challenges have you faced in maintaining your reflective practice?
How have you overcome them?
How do you make time for your reflective practice in your daily routine?
What benefits have you experienced from your reflective practice, both personally and professionally?
Can you share any insights or discoveries you’ve made about yourself through your reflective practice?
Have you noticed any patterns or trends in your thoughts and feelings over time through your reflective practice?
How has your reflective practice influenced your decisionmaking and goalsetting processes?
Have you ever shared your reflective practice with others, and if so, what has been their reaction or feedback?
What tools or resources have you found most helpful in maintaining your reflective practice?
What advice would you give to someone who is considering starting a reflective practice like lifelogging or journaling?
How do you measure your progress or success in your reflective practice?
Are there any specific events or experiences that have significantly impacted your reflective practice?
What role does technology play in your reflective practice, and how has it influenced your process?
In what ways has your reflective practice contributed to your overall wellbeing and personal growth?
Describing Inner Experience?
What am I trying to do today?
What is the future of computing?
What are the frictions that exist within short term and long term memory?
Is there any research on the effect of cognitive load theory on metacognitive processes?
Do you think it’s important to be doing something that gives this deep satisfaction?
What concepts do they mention a lot together?
Are there concepts that seem to be used similarly for different people in the same context?
Square = Triangle?
How does a journaling session end?
Can he explain specific anecdotes where he felt like something happened?
Does this stuff make sense to someone who hasn’t been thinking about it like I do and doesn’t have the same prerequisite context?
How can potential cognitive biases be identified and mitigated during the sensemaking process, as they may impact the interpretation of information and the decisionmaking process?
What strategies can be employed to maintain motivation and engagement throughout the sensemaking process, particularly when dealing with complex, ambiguous, or emotionallycharged situations?
When do they transition into a more reflective state?
What makes a shower so effective?
What neurophysiological changes are happening when you perform various activities?
How does that support the steps within the activity?
And the subsequent activities after?
Depending on the type of session, maybe different preprompting?
How do we search for representations, discover residue, instantiate representations, maintain representations and ultimately do the task?
What do people do when they want to understand stuff?
Study: how do people make sense of a body of work in order to answer specific questions about that work?
Where does the number 10 fit in this sequence?
What happens when we don’t need results instantaneously?
What other ways to search can we employ?
How do people accomplish this goal today?
What have we tried that could help with these two problems?
Why do we create abstractions?
Can we discover and explain highlevel patterns?
Do they generate insights that guide our design?
Do they spark ideas?
Do they lead to great systems?
What type of abstractions do I find myself creating?
What is difficult about creating abstractions?
Grab all the salient items, put them in one place, then ask what makes these things related?
what are they trying to tell me?
So where have I been getting stuck?
What was the problem that it solves for?
Why do people take the action?
Importance scale why do I care about this concept?
Why do I feel this concept is important?
Why might this concept not be important?
What do I know about this concept?
What do I want to know about this concept?
What are you pivoting on?
Why might this concept not be important?
What are people trying to do?
How do they do it today?
What are the challenges and struggles?
What emotions and feelings do you want to amplify?
Like, what is journaling?
Like what does it mean to get insights out of it?
Like, what is hard about that?
Um, Where do people struggle?
Uh, what do really effective journalists do?
When people do succeed with journaling, what is it that they’re doing?
Um, what are other activities that are like journaling?
People do instead, like, is this subsidiary analog?
How would my past self have responded to this prompt?
How do these themes compose?
Or do you have themes referencing the source text at different levels of fidelity?
Have you also considered doing some research and reading up on the topic of entrepreneurship?
What makes questions useful?
And when?
Am I eating well?
Exercising?
What would it be like if the entire interface were voice based?
How would you navigate while making commands?
Ring (has intuition on something you might want to hear?
A video game that you are constructing on the fly?
What are they?
What questions am I trying to answer at each time step?
Are there questions that I can’t even imagine to ask?
What might these look like?
How can I get at them?
Also, what role does text play in our lives?
Does the processing all happen at a more emotional level?
How will this make you feel?
Can you get someone to feel an emotion that they’ve never felt before?
What does san francisco mean to you?
if I just paste a url with no context, how come other people might know why I shared it but a computer might not?
what is this entire interconnectivity conversation all about?
I want a relationship, what is the content?
relationship, for what?
what if I want to have this kind of lifestyle and then go from there?
film looking at the snow scene, nothing else, what am I looking at?
How does language play a role in reflection, meaning making?
Want to explore the types of themes that GPT can extract from a text?
- has anybody actually tried any of this stuff?
Is there an expert model for this?
What are the downstream consequences of this hypothesis being correct?
Really important goal for the project is to understand whether anyone has tried anything like this and what do we know?
Do LMs continue to improve, eventually encapsulating KBs?
Or do they find some way to evolve in parallel?
What do they like?
What else does that imply they may like?
What do they know?
Where might they be confused?
What am I doing with this prototype?
What cognitive activities am I arriving at?
What if you could paint the text based on various attributes?
Embeddings of the lists?
Search for passages based on list embeddings over passage embedding?
Why would you want this?
How do you get the system to justify and then reweight?
Does this work on journals?
Does this work on something authored for someone else?
Commonsense reasoning?
Hierarchies?
What is the author’s main idea about the intersection of agents, LLMs, and programmable attention, and how does it relate to transformative implications?
How does the proposed agentenabled by LLMs, observe an expert in an environment, learn to mimic the ways the expert attends to that environment, and how can novices use this agent to achieve similar goals in that environment?
What are the challenges faced by the author in understanding the construction of common sense reasoning structures and their importance in making creative leaps?
How can the author overcome these challenges?
How does the author perceive the connection between concepts in order to spark creative insights, and what are the underlying patterns and connections that lead to these insights?
What are the distinct knowledge bases and how do they draw from different sources to reach their own conceptions of which relationships to use?
How do they simplify the graph and create more continuity?
How does the author use concept mapping to guide the user with questions, and what is the system one that the author proposes?
How does this system help in solving problems and making decisions?
What are the hierarchical clusters of concepts, and how are they similar to fire together?
How does the author use these clusters to guide the user in solving problems and making decisions?
How might the design of an agent enabled by LLMs that observes an expert in an environment potentially transform novice learning in that environment?
In what ways might the construction of common sense reasoning structures be important for making creative leaps?
How could a concept mapping system that structures concepts in a time dimension and creates hierarchical clusters be designed to guide users in reflecting on past experiences and finding solutions to current problems?
What are some potential challenges in creating a knowledge base that draws from multiple sources and aims to find universality in verb structures?
How might a system that mimics the way the brain follows patterns and connections between concepts in order to spark creative insights be designed?
Can you summarize the main idea of the text?
How well does the author support their ideas with evidence and examples?
What are some potential strengths and weaknesses of the author’s proposed system for concept mapping?
Can you identify any gaps in the author’s argument or areas where more research is needed?
What are some potential implications of the author’s ideas for the field of artificial intelligence or education?
How well does the author communicate their ideas and concepts in the text?
Can you identify any assumptions or biases that the author might have?
How does the author’s proposed system for concept mapping compare to other existing systems?
How feasible do you think it would be to implement the author’s proposed system in practice?
Can you suggest any alternative approaches or solutions to the problem that the author is addressing?
Can you provide an example of how the latticework approach can be applied to connect seemingly unrelated concepts?
How might using distinct sources and tuples lead to a deeper understanding of complex information?
How can writing in different environments and drawing on diverse background materials help improve the clarity and effectiveness of our writing?
What realworld examples can we use to understand complex concepts such as love or morality?
How might a person’s emotional state affect their understanding of a particular concept?
Are there any examples of concepts that are universally understood across different cultures and societies?
How might the absence of narratives impact our understanding of the world and our ability to communicate ideas effectively?
Ok what properties does biking have?
What things have similar properties?
Can you observe the relational structure of someone deemed creative?
What semantic labeled edges does it follow?
And how many steps does it take?
Can you measure creativity through the generated semantic graph?
How does this connect to the projects we’ve (me and Andy) discussed?
What are the lesser known strains of AI research that may become really influential towards “AGI”?
How do augment LLMs with a reliable, flexible and powerful long term memory storage?
Are the contexts the same?
Some people do this much more effectively than others, what are they doing?
And can we allow more people to do that kind of thing?
How do you even observe this?
Difficult to enact?
how to get this info?
What is creativity?
What is learning?
If we find the right prompts and express them at the right time?
Can you arrange the environment or context to, to make those things more possible?
if I actively analyze the connection between X and Y, does that make you more likely to notice these subtle patterns and connections?
Really important goal for the project is to understand whether anyone has tried anything like this and what do we know?
Is this how David did Midjourney?
Because some of the functionality literally can’t be done without computers?
What are the downstream consequences of these things being correct?
How Do Artists Differ From Bank Officers?
What is this opportunity space?
Does oblique strategies help?
Does using Bloom’s taxonomy help?
How do these things make people feel?
what’s the expert model?
think about some practical questions about what domain to try that?
has anybody actually tried any of this stuff?
Is there an expert model for this?
Has anyone tried an intervention based on the patterns he’s revealed?
Why so many dysfunctional institutions?
Ask yourself, is there a better way?
Would you respect someone else for taking the lead on this?
How you move things from the former to the latter?
What if to favorite something, you have to engage with it in deeper and deeper ways?
In what contexts/scenarios/environments do people already manage to assimilate these sorts of things?
What does the literature already “know” about assimilation of this kind?
What seem to be the key barriers to assimilation of this kind?
So here’s one more vectoring style question: is there a particular domain where you’re especially excited to investigate this kind of assimilation?
How will this affect people?
How do go about assimilating new representations, concepts, processes and techniques today?
How we can learn to assimilate new representations, concepts, processes and techniques into our capability infrastructure?
Where in life do you feel you have the MOST opportunity?
That is, where in life do you have the greatest freedom to do what you want or the most effective skills that enable you to modify circumstances for the better?
How are you feeling?
In what ways do you think this belief has harmed your ability to see?
Do you want to use it?
What do certain representation lose?
What do they gain?
What does arithmetic do?
Why?
How?
How we can learn to assimilate new representations, concepts, processes and techniques into our capability infrastructure?
How do we usually go about this?
What if we used ideas from spaced repetition?
What type of general principles did the researchers learn across disciplines?
And how were there principles instantiated in specific cases?
Is this because they are independent enough from daily activity that you can find true novices that can’t easily draw from domain expertise in other fields?
How do we engage with the material of a text over time to internalize it at a deeper level?
How much compression can you get in a single word?
How are their perspectives different?
What are some of the themes of Cedric Chin’s work?
What if the you can train the agent yourself?
Possible exploration: What if we constrained the agent to only observing actions from one main application (ex: google docs) and focused on observing the actions an expert makes in the process of providing feedback to someone else?
What if the novice acts as both teacher and learner?
But how much does this activity change how you engage in the world?
What if all that theory stuff isn’t important until you have built experience taking these little actions bottoms up?
What makes a conversation work?
Do you remember the insights that were crucial in learning to ride a bike or drive?
How about the details and insights you have that led you to be good at the things you’re good at?
what things have I already learned in the past?
What has become intuitive?
Might want to ask why?
Why is this activity useful?
How might an expert writing instructor attend to a piece of writing?
What would they notice and why would they notice it?
why do you move chunks around?
What are you trying to say by this line?
What is the idea here?
Can you explain what your definition is of this thing?
What base affordances are needed by the agent to enable such activity?
What are some of the assumptions?
Operating on the whole text, compress as necessary, like in a conversation with someone and you’re like what did we just talk about?
How would you do an exercise like the one we did for the NYC writing club?
What have you learned collaborationwise over the past few years?
Making music, or can there be a conversational equivalent?
Exercises?
Is there a latent genre that we don’t understand, that would give us some insight into the tools we need?
As a writer, when this person says they violated the agreement, are they saying the same things?
Throw this into ChatGPT, can you give me an outline?
How do you do it?
Are there things I’m not considering?
Frames I could adopt that would change this feeling?
So what did you do today?
How did that connect to your goals?
“Always grab your audience” what should the audience seek next?
What other verbs could be useful for nonmemory prompts like these?
Salience over what?
Do I seem angry?
angry over what?
Which ones are mentioned most often in association with each other?
What’s the simplest version of references I can think of?
Where to put that list?
Beneath or above (depending on screen space) the chunk of text?
What information would be useful to show in a reference modal?
But what if you want to emphasize one section?
But what if the components of a theme are spread across multiple sections?
is it better that these are all aggregated in a single place?
Or distributed in sections across the timeline?
What was the question that even motivated this project in the first place?
Who ignited the movement?
folding?
What is interesting?
What might support or contrast with this particular concept or point I’m trying to make?
where else is this card referenced in this space?
or in other spaces?
The map shows what?
What areas are similar?
What areas are different?
What areas are visited most often?
What is the smallest set of interface elements that can get the intention I want?
What are my intents?
Filter how do you find all the good stuff?
Ok what am I writing about?
Why did I arrive at references?
Often mentioned?
Frequently mentioned?
what are the jobs to be done that these tools support?
What other next steps or moves seem worth exploring next to you?
Often mentioned?
Frequently mentioned?
What parts do you copy over?
A piece of a passage?
How do we identify which references are powerful?
Which are mentioned often?
Which ones we returned to most?
A forum + commonplace book + references had a baby?
what about labeled edges?
What did the slip box enable?
What’s the point of that?
What is knowledge?
What did the slip box enable?
What parts do you copy over?
A piece of a passage?
How do we identify which references are powerful?
Which are mentioned often?
Which ones we returned to most?
What if user interfaces were crafted with the same type of consideration and thought that goes into a locally sourced meal?
Can food learn from interfaces?
What’s the equivalent for a really custom workflow?
What if semantic similarity can work?
But you have to treat it as a game of tetris?
What are the moves we can make to trigger new thoughts?
What is the hard thing about making sense of brain dumps or stream of consciousness writing?
Like if we created some intervention, designed some interface, enabled this type of action to be taken, how would we know that things have improved?
What are we comparing that intervention against?
What is the strongest counterargument?
Like, what is journaling?
Like what does it mean to get insights out of it?
Like, what is hard about that?
Um, Where do people struggle?
Uh, what do really effective journalists do?
When people do succeed with journaling, what is it that they’re doing?
Um, what are other activities that are like journaling?
People do instead, like, is this subsidiary analog?
Why do I want append only?
Hopefully you’ll produce a real increment, and you’ll build skills, and that’s great—but who’s going to do the next iteration?
How can we develop transformative tools for thought?
how can we develop tools that change and expand the range of thoughts human beings can think?
One of the most famous papers in the philosophy of consciousness is entitled “What is it like to be a bat?” Each tool for thought poses a similar question, near impossible to answer without immersion in the tool: “What is it like to be a language user?
So what?
Would that mean I shouldn’t do it?
How might one create timeful texts—texts with affordances extending the authored experience over weeks and months, texts which continue the conversation with the reader as they slowly integrate those ideas into their lives?
What other powerful tools might it be possible to create, making future books into artifacts that transcend their pages, as they slowly help readers shape their lives?
What if the thing we want to be emphasizing is only mentioned once?
Like was it a quote from a text?
Ok so with that out of the way, what do I want to bring to the meeting today?
If the key point is append only, what are the consequences?
Why did you do such and such action?
What were you trying to figure out?
What resources did you access and for what?
Like can you take this section and let me know if anything like this exists?
Let me know if there are things that look like threads I’ve opened up that I might want to continue exploring?
Table of contents, what if it went backwards and post box created the table of contents?
why did you start exploring this question?
What might be a counter argument?
What am I trying to do here?
How did I get to where I am now?
Many to many?
What matters?
Why does this stuff matter?
What problems does the interface solve?
I bring that theme into the working context and ask, where else does this theme exist?
What does that look like for text?
What if instead of limiting things to a few primitives, you opened the floodgates and allowed for any type of prompt manipulations?
Is there a difference between raw material and theme pieces?
Can you just treat them as similar?
how do you want someone to feel?
What are the consequences of reframing everything as such?
It could be a vague question like what am I doing?
What if you have a really basic HUD?
Like the equivalent of a persistent text box to leave a note of what the big picture goal is?
Is there a latent genre that we don’t understand, that would give us some insight into the tools we need?
What are the calculations that we are doing?
What if this system is just letting you be lazy?
How do you balance these two things?
What if I made that the focus?
How would that change things?
But what does generalize?
What parts are vague?
That I’m struggling to articulate?
What do I want to use?
What friction are you solving for?
Why are trains so nice?
In particular, if there really are 8 (or 12) themes, will the colors suffice to distinguish them?
Or will the emoji have to be more proximal?
Or… patterns?
Multiple colors extracted from the emoji?
What if there was something similar with interfaces, systematically reintroducing elements back?
What if those elements could be things that didn’t exist beforehand?
Does this shape work to solve my needs?
Does it do so elegantly?
With information, is it possible to understand this question better by organizing the elements and their relations in a way that emphasizes the important parts better?
Who ignited the movement?
Don’t tell me what you think others will find interesting or will understand, what do you find interesting?
Is there a connection to life as improv?
Maybe for human ones too?
What are we trying to do?
So for this week, I asked, what happens if I focus on just that one piece and explore the move space around that activity of extracting themes?
What if I focused on that one piece as the core job to be done?
What is it like to be the one making this ethical decision?
Do I feel like I’m trying to get away with something?
“theme”: “Positive moral obligations for AI safety?
If utilitarians decide that the benefit to society from economic productivity justifies the suffering from job displacement, how will we as a society help those harmed?
What happens if I make something an exception?
Where does it make itself prominent?
In relation to what other themes?
What does that look like for text?
What do I want to do this coming week?
“tell the data’s story” Is the aim really to “tell the data’s story”?
Tell it for what purpose?
To whom?
Naively, given our chats the last couple of weeks, I think I’d have written “…features and patterns to figure out what to do next.” Was there a reason you moved from the “what to do next” framing we’d been discussing to “telling a story”?
We’ve been talking about it plenty in our conversations, but one of the next things this statement will need is: what is the point of this labeling, connecting, and categorizing?
the first sentence)?
What do people do with these labels, and why is it a problem that they get stuck?
It’ll be interesting to discuss: in what ways are these scenarios (drawing from objective “data”) like, and unlike, the scenarios we’ve been discussing (drawing from messy qualitative “kindling)?
What does software wish to be?
Structure and art, useful distinction in writers?
What is a website for?
First there’s the question of what is creativity?
What do I care about?
Why am I doing what I’m doing?
What are the affordances of physical space for organizing your thoughts?
What are the challenges in physical space for organizing your thoughts?
What are the affordances of digital spaces for organizing your thoughts?
How would it feel being in that space each day?
What types of artifacts would we want to produce?
What problems are you trying to solve?
What types of things do we want to create?
How can we build a powerful pipeline for creation?
What affordances does it offer?
What type of structures can you create with 2d space?
When moving to physical, what do you lose?
Can you bring it back?
What does an information map want to be?
Help who?
Help when?
Help where?
One infinite canvas?
Or adjacent canvases like in Figma?
The workspace, what is it?
What is the equivalent in a like steampunk world?
What would be interesting to talk about?
What parts feel most important?
What would be interesting to talk about?
What parts feel most important?
What is a block?
What are the precise actions and transformations that need to happen to the content?
What are the mental tasks involved in synthesizing meaning from source material?
How will the person manipulate and structure the concepts?
It’s asking what is most important from this material to carry with you?
To nurture and develop and review?
And unlike the burnouts I’ve experienced before I find myself asking the question: “so what?” Why is working so hard important?
I guess a question I’m pondering now is what mechanisms can be introduced into a curation tool to create this feeling of aliveness?
Why do I feel like this?
For each day that mentions depression, or some other peak mental state, are there any patterns in the series of events that led up to that?
what is the cascades for an expert reading a paper?
What do they attend to?
How do we weave a cohesive narrative around the massive amounts of data, separate media silos and new interaction models?
Imagine being able to explore interactive life stories of people?
See patterns and themes: identify clusters — does periods of intense activity indicate something was about to happen?
Notes or stories can be manually added: what was in my mind when I was writing a particular poem?
Why was this period particularly important to me, even though it seems like a string of random events?
Can we learn who a person is and what they might like from a few high dimensional data points?
What would X say?
Can someone be motivated to contribute simply from ability to learn?
What motivates someone on reddit?
Where do you see yourself in the next 5 years?
What if we could use the algorithms to understand ourselves better?
What if each person had a tool that could serve as a guide through the complexities of life?
Who is the equivalent of X in Y?
How was this realized in society?
What struggles did it have?
What parts of previous computing pioneer’s work am I drawing from?
What is the lens that we use to introspect?
How do we become aware of these things with time?
Why now?
What specific types of inquiry does commonsense reasoning + computational language (wolfram alpha) enable that general purpose language models struggle to do?
What makes a good conversation?
Can a system return useful long range connections to the thoughts someone is having in the present context?
How do people respond differently to suggestions made by a person vs. computer?
Do people even want advice?
What makes someone receptive to advice vs. not?
What is the role of memory decay and is that a good thing to have in a system?
What if we could externalize the algorithm’s understanding of us so we can inspect them and perhaps better understand ourselves?
If the work I did succeeded, what is the transformative potential for such work?
How would it be a more desirable world to live in than the one we live within today?
If Google can understand us, why can’t we use those algorithms to understand ourselves better?
How did this design exercise go?
What did you like and dislike?
What could have gone better?
How did you enjoy your day?
How was your conversation with Patty?
Was today any different?
How do you think Patty’s worldview compares to yours?
Was there anything surprising?
How are you feeling today?
What would you like to feel?
Do you personify the agent?
Could you get an AI to perceive the connection between byung chul han and bret victor?
It should ask a question like, do you know of this person?
Then the followup would be why?
What are the limitations of embedding space?
How are they related?
Like hiking?
Why am I doing what I’m doing?
Are they internal motivators or external?
Expectations on me?
What would be the implications from these ideas as Moore’s Law gave us increasing and predictable more compute power?
What might be the implications when these ideas started to converge with one another in a new world where everything could be coded in digital bits?
what if you attempted to get AI to extrapolate key features of a poem?
Can you solve ontology problems with probabilistic models on top of formal logic frameworks?
can this help us gather long range relationships in data?
Intuitive of metaphysics, all that stuff, if you were to create a website that described intuitive physics into the psychology, would it be bigger or smaller than Wikipedia?
What do you think?
What went wrong with these types of programs?
Why have they not succeeded?
How should I solve this problem?
What is the place of question taxonomies?
Maybe you should try this related phrase?
What stuff does it contain?
Does it have intention, questions?
Is it reflective?
brainstorming?
topics or themes?
what does it lack?
when would the material become useful again?
Are there any sections where it might be useful to reference?
Or are they the same?
Can you build a similar machine attention mechanism for conversation?
Can this system just be a full speech system?
Talking and having a dialectical conversation?
If I don’t understand it, how can I say I do research related to AI?
can a model reify life trajectories, resemblances to past selves, and do both the boring and unexpected?
What if you removed the conception of H tags but kept their ability to create hierarchy?
Why only the questions?
What is context?
Can we develop a rich context, structured data from unstructured information?
If IFS is true, what is the minimal encoding?
Have zoomable user interfaces reached their final form yet?
Can computers get to the semantics of language?
how can computing as a medium support how we think?
And how might is map to the ways we think today?
How do we humans think?
What would viewing a paper in the computational age look like?
How can we allow people to see in new ways?
What is ChatGPT Doing… and Why Does it Work?
Why now?
What can I learn about the inner workings of machine attention mechanisms and what might this suggest about designing attention mechanisms aimed at supporting human cognition?
When are these various column actions actually useful?
Is there too much overlap in some?
Can you identify when a certain action would be useful based on some attention mechanism?
how to make attention more dynamic?
As simple as keyvalue store mapping to various parts of memory that it can retrieve for the task at hand?
Even better than specifying the pattern, can you just give GPT3 the examples you like and then have it extrapolate the regex pattern that would capture similar things?
How do we make that a part of people’s process?
what if what we need is a different interface for these types of cognitive activity?
Ones that are finely tailored to each of these forms of cognition?
What is actually happening during a good conversation?
Which is better?
What if you could segment various logs based on the activity?
Can you track this change over time?
Notice a change in the conversational dynamics and shift the flavor of questions that are being served from the system?
What is the minimal app, data structure or system for that process?
Daily intentions, pull up some of the old things, are any of these things salient for you today?
Is it actually important to you?
What is preventing you from making progress on this?
Why are you struggling to complete this task?
Is there something deeper to unravel here?
What happens if you think about a software company like this?
Why were you interested in something?
How did you discover something?
What did you take from the piece?
(Summary) What do you want to do given the piece?
But what does this suggest for others?
What are my priorities today?
What is the actual time that an event takes (travel time to and from)?
What feels like the final frontiers of exploration?
What if many aspects of cognition are unnecessary?
What specifically are you able to do with novel datasets?
Given the premises, what do you think about this?
What is the workshop for?
Is what I have for research support tools (and in a way, creativity support tools) sufficient?
What is the challenge with writing?
What happens if you push LLM use to the limit for a single individual?
How could you give new superpowers to people from pushing LLMs to their limit?
What if you wrapped partial habits in a daily habit?
Aggregated them?
What does that look like for text?
What happens if I make something an exception?
Where does it make itself prominent?
In relation to what other themes?
Why is it better to compare the frames of text rather than the raw text?
How are themes used?
What do I want to do this coming week?
We can even apply the technologist’s process to technology itself, answering and acting on the questions: how best do we make reason out of and model the madness that is technology?
What is it like to be the one making this ethical decision?
Do I feel like I’m trying to get away with something?
“theme”: “Positive moral obligations for AI safety?
In that world, how do we coordinate as a society around an ethical framework?
How do we ground morality in real situations?
What if instead of limiting things to a few primitives, you opened the floodgates and allowed for any type of prompt manipulations?
What was the question that even motivated this project in the first place?
What led to the increasing feeling of disappointment in the early middle weeks of the research?
how do you want someone to feel?
What actions would allow these feelings?
Can you extract the key themes mentioned in the text above?
For each theme, can you quote the sections of the original text that support it?
Can you take a braindump and section it up into reasonable headings?
Or can you take a braindump and and just pull out the themes?
Between “Beginning of Janurary” and “Week 1”, where are there continuities in the themes being explored?
How about discontinuities?
Three levels of headers?
If it were so very easy to look things up, how would our vocabulary develop, how would our habits of exploring the intellectual domains of others shift, how might the sophistication of practical organization mature (if each person can so quickly and easily look up applicable rules), how would our education system change to take advantage of this new external symbolmanipulation capability of students and teachers (and administrators)?
What does this look like for language models and increasingly cheap reasoning?
And what exactly is the audience seeking that reads this beginning?
What’s actually going on?
For what class of problems does paper and pencil help?
For what class of problems does it not help (or hinder)?
How much can it help?
What if instead of human versus machine we played as partners?
Can we find principles that will help us design more powerful humanmachine systems?
What’s missing from Engelbart’s paper?
Can we trust the computer’s ability to help us identify key parts of a text?
Will this just be eclipsed by machines in near future?
In order to.. what?
Feel safe?
Feel worthwhile?
Feel acknowledged?
Feel successful?
Feel fame?
What if you feel nothing?
How would that be?
What if you let go of who you thought you were, and figured out the truth?
What if what you thought about these emotions weren’t the full picture?
Would that change everything?
Can you help me better understand Minsky’s beliefs around the above concepts?
Can you build on your understanding of uniframing, accumulating, reformulating, and transframing and incorporate the above ideas (memory stuff) from Minsky into this framework?
Are there other ideas from Minsky that we can incorporate into our understanding above?
What parts does it cover and what parts does it not?
How do we deal with the complexity of information?
Unified theory?
How about other frames that Klein suggests like stories?
What is each more effective at goalswise?
What parts don’t fit?
Handling new evidence, how does it affect the models?
Because how could I waste a CEO’s time with the sorts of petty projects I was working on?
And if that’s the case, can we see the humanity in ourselves?
Can we love and see each other for what we are?
These complex beings living complex lives in a complex world with complex situations?
How much do I want to just introduce problems without a solution?
Why do you keep coming?
A father and son asking the question, How do you preserve the impossible to preserve?
Or, in other words, how do you stop time and death?
What makes a life?
Roughly, what could I imagine the constraints to look like?
And if that’s the case, then can I allow myself to let this go for now?
Focus myself fully on the task of discovering what is possible?
What does that look like?
How can I shift my mindset from focusing on feature sets to truly embracing a researchdriven approach, seeking powerful ideas and solutions?
How can I let go of my preconceived notions about my future and stay open to radically different possibilities, while fully immersing myself in the present task and honing my skills, taste, and communication abilities?
How do we get to a really powerful explanation of what is happening?
How do we make sense of this?
How can we make this fun?
Can we change the resistance that people feel to it?
Had did Deutsch come up with new explanations?
should I consider moving some of the global state into the objects?
What does it look like to generate new meanings?
How far can you take this one mechanism?
Of pointing, selecting?
How would that look?
What else is on my mind?
Why?
But why?
And is there a place where confrontation could be useful?
What was I excited about?
What was surprising?
How might we inspire or support {motivation, interest, curiosity, connection} with a topic?
Could LLMs help make reviews more effective or meaningful?
Perhaps you might want to have a conversation about a prompt as an exercise?
Could you situate a memory system in the context of use somehow?
Could these memory reviews be situated in a social context?
What incentives do scientist have to make their data public and or partially open?
How does Data driven intelligence relate to the tools that amplify collective intelligence?
What’s important?
Throw this into ChatGPT, can you give me an outline?
Like how does one feel like they’ve made sense of a set of questions sufficiently enough to close the narrative?
I guess when the question feels answered?
And then what is the form of the final output?
My instinct would be to focus our conversation on this notion of representations: what kinds of immediate representations end up getting produced as they work and think?
Are they always aggregates/compressions, or are they sometimes something else?
How do the aggregates/compressions interact with “raw data”?
How do they end up using those representations to produce some “final product”?
Are there techniques that you use now that you didn’t use at the start?
But what is the unique lense or frame that he is bringing to the table?
Many ways you realize you have the outline the wrong way?
As a writer, when this person says they violated the agreement, are they saying the same things?
Trying to make an outline, how do you do it?
How do you do it?
Memorization and Piotr Wozniak, what did the psychologists tell use about memorization?
Is there a latent genre that we don’t understand, that would give us some insight into the tools we need?
Did I record this or not?
Ok where do I go now?
Do I find what I’m looking for or something different?
Can you pull out all the quotes from the text above that may be useful to this interest?
But is thinking actually that linear?
Why do I let these people have that type of power over me?
Is it because you value their opinion or because it touched on an insecurity?
What can you take away from this situation to help you become more resilient in the future?
How could a company have that long for development of just a single part of the game?
Ok how am I feeling otherwise?
Open question, how mutable are these objects once created?
Human attention and its role?
With a particular focus to what makes them hard?
What makes a good concept?
Are penguins birds?
Is the pope a bachelor?
Why do we need concepts and how do we use them?
Ex: Why do we need concepts?
[00:00:11] Matt: in your wall?
How are you doing?
W were were, were there like a theme that you were sketching around together?
Cuz I, I can get all consumed to something, but like, I have to be very explicit about like, you know what, I should go and like not do anything for like the next like, Afternoon or whatever, right?
[00:02:50] Um, What, what is hard?
Like why is it so hard to get into an ingroup?
It’s like, what’s hard about that?
Uh, so, um, what, what would it be helpful to discuss?
Like more like linguistics type stuff around like what is the point of conceptual abstractions?
Um, starting with like, why do we even create these abstractions in the first place?
It’s like, okay, so we’ve identified this pattern, uh, what are we gonna do about it next?
Or like, also, I guess this works for you with yourself over time, right?
[00:08:57] So, uh, for example, why does this concept feel important?
Why might this concept not be important, um, in reference to some sort of objective, like some goal, um, What do I know about this concept and what do I wanna know more about this concept?
So this next question, what type of abstractions do I find myself creating?
[00:12:58] I, I guess now I found myself wondering like, how is it different from a highlight?
Is a, is a waypoint like a, uh, range list highlight?
Of zero?
What is a pivot?
Um, yeah, I found, for example, one thing that I extracted from our conversation was, um, like, what is the problem with working with these brain dumps?
Uh, oh, you mean like the phrase is the abstraction here?
Like just keeping this in mind, like what is the problem?
Um, those, those are, those are like contexts, right?
What do you mean by angle?
Then, um, is it that we’re trying to find one noun that captures all these things?
And one question I, I guess I remember asking last time, uh, on Monday was like, well, if this is a thing that people do already, um, how exactly is it that we imagine we’re helping?
[00:25:00] Um, one thing I’m curious about, maybe if in your own experience, um, is it like, like arriving at these sort of clear like phrases or like terms, et cetera?
Like if, if you sort of like orient this writing process around that, is that, like, would that be something that maybe like is more of like an expert practice and if, if so, would that be something you would want an interface to nudge you towards?
But you asked, um, would it be helpful to have a piece of software to nudge me towards that?
To nudge me towards that?
I, I think we’ve, or, uh, on Monday I had with like, um, gathering sort of those candidate phrases of like, uh, at least yours, right?
Like, okay, like, are, are we doing this?
So I mean, I, I guess like on Monday you heard me recite like, okay, so we want to figure out like what is actually the problem we’re solving in what context?
[00:30:39] Or like, why is it hard, uh, or like, uh, Uh, uh, what kind of moves can one make just conceptually, uh, in order to like, dig oneself out of this problem?
Um, what is like emotionally salient as I’m going through it?
I think possibly to an answer to a question I asked earlier in the conversation, which is like, what is it that we imagine we might be contributing here?
Like what did you do?
[00:38:01] What does it let you do?
How, how, how does it, how does it solve your, your problem?
Like I think you’re saying you see this label like research logbook angle and you don’t really know like, what does that mean?
What do I know about it and what do I wanna know more about?
[00:43:03] A highlights interface in PDF apps is usually, like, there’s a list view where like all of the highlights are like listed in a table and you can like click one to navigate to it, you know?
Um, now you need to like look at those things you’ve extracted and try to figure out like what’s next or like the so what or, um, like, uh, what, what are you, what are you gonna be able to do with this intellectually now that you’ve extracted all these things?
It’s something to do with like the just like, uh, the spatiality, right?
[00:47:19] So do, like, why, why am I so mad about this?
Is that what you’re asking?
Uh, why am I so mad about this?
But I mean, so like, I’m pasting these large chunks of text because if I highlight only like the, the most interesting part, then I will end up with even more garbage notes, right?
Like this was a book, right?
This is structured text, right?
Really?
[00:56:37] Does that feel better?
No, what’s that?
Um, do, do you know this one?
What was that?
That is the nature of the problem, right?
[01:13:33] Matt: days?
For example, are they typically native or nonnative?
What about the type of bark?
What method do they use to reproduce?
Are they seed or fruit bearing?
how do we train people to come up with things that are completely new?
Why study history if history never repeats itself in Illstructured domains?
Why are the moomins so popular 70 years later?
20 years after her death?
DONE: Will start with this article and delve further if needed Is Speed Chess Good For You?
Why do we have memory systems like what what is memory for and what do we care about, right?
what type of change do I want my own lifestyle?
And I think a lot of people are trying to figure this out like, oh, what’s the point?
I can add a title I can add like a status like what action like what have I done with it?
Tags basic stuff, impressions, like, what was my initial impression when I first added it to the table?
Why did I add it?
What was the thing the question etcetera, that I was hoping that this piece of content would answer for me, I found via table tab or column and this is just saying like, where did where did I find it on the internet?
Or how did I come become aware of it?
And I have to go through multiple passes of sort of writing and then like trying to extract like the key bullet points and like, okay, are these bullet points good?
Play around with the order?
You like there’s a number of heuristics don’t have orphan cards is one so you know, if you’re going to just like get this one facts here, or this one other facts here for a narrative, maybe that’s not something that would be very useful or effective to really try to memorize but the other bit too, is is this is is encoding this piece of information in my like memory, is it going to be useful?
For you know, is it going to stay five to 10 minutes in the future?
I guess one of those pieces of feedback that I’m sort of reflecting on now is, Who am I trying to write to?
And this light, this like device that like this night lamp that this company Casper had designed?
And I think in that you can see like almost like the love for people and that like these designers have when they’re considering what what does this person’s life look like and how can I design something to fit within their live their life and their lifestyle and their activity?
Zooming out what is the like kind of big point that I’m trying to make with this?
This piece?
And yeah, I guess this also kind of touches on like, Okay, so who is the audience like?
I’m writing I’m learning what is it kind of accumulating too?
Like I’m like, Why?
Why do we have memory systems like what what is memory for and what do we care about, right?
And so out of those things, like if I look at my own lifestyle, what type of behavior modification what type of change do I want my own lifestyle?
And I think if like me like yeah, change, I wanted my own lifestyle, but also how can computers and technology etc, and behavior changes kind of like support that?
And I think a lot of people are trying to figure this out like, oh, what’s the point?
All this stuff right?
so then, I guess the question that I need to explore a little bit more with Andy this week is like what is the purpose of memory?
What is the purpose of memory systems like, what are we actually trying to get to?
But then from there, the question is, what type of experiments then?
And such Can you perform around these values and its desired possible end state?
What are the activities that we can do?
How does it feel to answer the same prompt over time?
How does asking follow up questions feel?
Why?
Why?
How can I get GPT to ask really salient questions to a person based on all the context I can pass to it?
Impression: How do experts deal with novelty?
How do we navigate fuzzy realms without much empirical grounding?
Is Speed Chess Good For You?
What happens as an intermediate player without experience playing speed chess?
INQUIRY: what if you went multimodal?
Embeddings?
Maybe layer on a music player interface on top?
What if you use a slider to sort slice through the sequence at different levels of fidelity?
Other pattern matching algorithms?
How did he ask questions?
What did they do?
e.g., ’Why do you say that?
e.g., ’Is this always the case?
e.g., ’Why do you say that?
e.g., ’What is the counterargument?
e.g., ’But if…happened, what else would result?
e.g., ’Why do you think that I asked that question?
’, ’Why was that question important?
Revealing the issue: ’What evidence supports this idea?
Conceiving reasonable alternatives: ’What might be another explanation or viewpoint of the situation?
Evaluate those consequences: ’What’s the effect of thinking or believing this?
let newAnimatedUntilTime = animatedUntilTime ??
For while loop swap, don’t you want to only update destination position so that animation step can catch it and perform spring animation?
Question first posed in How can we develop transformative tools for thought?
how do you pull that out?
how do you enable talent to do the things they want to do?
What is the point of computers?
Having progress indicators that counterbalance each other?
What does this entail for you?
up question: Why do you take interest in that person?
Why does this make you feel excited/anxious/fearful/hopeful?
but should these ever have a place?
Maybe push a single question as a starter?
What are you thinking about today?
What are you worried about today?
How are you feeling?
Send a question to X to answer?
How do we measure process in cognitive activies?
how do you pull that out?
how do you enable talent to do the things they want to do?
Asked, what would happen if in ten years you were told to stop using the device?
HowRU?
How do we give a sense of progress?
That we’re learning?
Go over what exists, what’s good about them?
What’s bad?
How do we show learning?
how do we see the forest vs. the trees?
If you review after several months and forget, does it mean you are making less progress?
Punch card form of progress instead?
More naturalistic forms of progress?
Could spaced repetition act as a puzzle piece to “unlock” new sections?
When is it good to reexplore a question?
How to cultivate curiosity?
Makes setbacks interesting, what can I learn from this?
Why does this seem important?
More powerful unifying theme?
This may be pulling out a proper noun or a noun and asking what does that noun mean to you?
What systems does this creative engine need in order to work?
A reflection/memory/creativity support system that in ten years, would be something people would be unwilling to put down?
What would it look like?
Most learning happens outside of educational institutions, how do we support that type of learning?
A lot of learning is downstream of cultivating curiosity and motivation, how to support this?
Why?
What are the main values and enablers I hope to convey?
What is the feeling I’m hoping to convert with the system?
I’m sitting in my single bedroom, shared apartment in NY wondering why am I doing this?
What am I trying to prove to the world right now?
To build what?
To say what to the world?
Why make this thing?
But is this whole endeavor frivolous?
What do I fear?
For if we didn’t love the time we get, why would we despair the end?
But what if we’re wrong?
When the war makes a more pronounced statement, will we be ready?
How would you solve this?
Why do you believe this?
How do you nurture these traits?
Ability to hold divergent views?
How do the nurturing curiosity and metrics of progress relate to each other?
What is the process under which this occurs?
Why do you believe something to be true?
How much depth do they have in understanding a concept?
What if you had a system that only asked you questions but could help you make sense of stuff?
How does X connect to the source material that had X as an action?
is there a physics thing for this?
Good collaborator?
What is a hard memory for you?
What is the worst memory for you?
Impression: How does attention mechanism fee back into weights?
Attention modifies its respective section of a layer?
How does attention mechanism fee back into weights?
Attention modifies its respective section of a layer?
What stuff does it contain?
Does it have intention, questions?
Is it reflective?
brainstorming?
topics or themes?
what does it lack?
when would the material become useful again?
Are there any sections where it might be useful to reference?
How do human attention and machine attention mechanisms compare and contrast?
Can you identify these powerful question sequences for people to use in their daily lives?
What can we learn from each type of externalization?
What type of analysis can we do on various forms?
How do you evaluate truth?
Does this suggest any general truths about becoming an expert in something?
DONE: Will start with this article and delve further if needed Is Speed Chess Good For You?
So what is it about speed chess in particular that makes it correlate so highly?
Does it map closest to the real world conditions?
Is Speed Chess Good For You?
does it suggest any general principles about becoming an expert in anything?
The need to practice intuition based decision making with limited information or analytical time?
Does this just require intense amounts of motivation and will?
Can making the practice more rewarding, interesting, fun play a role?
What are these?
What do they contain?
what are the categories in intelligence augmentation?
What can you do?
How to cultivate curiosity?
Curiosity makes setbacks interesting, what can I learn from this?
Why does this seem important?
Is the a better more powerful unifying theme?
What evidence can you point to to support your belief?
Why are you reading this and what are you hoping to learn?
What are you engaged in this activity/relationship etc?
For doing your research, why are you doing this?
Is this a strong argument?
Are the sources for this argument strong?
Why might the conclusion be wrong?
How does the author use the term X?
Is this different from your own definition?
What are the main categories of overlap between the two models?
What are my goals for learning?
Does this material make sense to me?
TERM: focuspluscontext view on the data?
What other analytics would be interesting to learn about the question set?
What is ChatGPT Doing… and Why Does it Work?
How can we evaluate meaningful outputs from these models?
Are there “semantic laws of motion”?
A geometry around how GPT moves around embedding space with each token generation?
What can I learn about the inner workings of machine attention mechanisms and what might this suggest about designing attention mechanisms aimed at supplementing human cognition?
How can we evaluate meaningful outputs from these models?
Are there “semantic laws of motion”?
A geometry around how GPT moves around embedding space with each token generation?
What can I learn about the inner workings of machine attention mechanisms and what might this suggest about designing attention mechanisms aimed at supporting human cognition?
What ideas are researchers drawing from human brain cognition to apply to neural network design and architecture?
As such, what happens if we had a completely different computer architecture that could explore this?
Who is the person who might be able to do this work?
Did their life look to be a completely different trajectory until someone discovered this problem formulation that mapped to their skills?
What tricks do networks like chatgpt use in order to make up for the lack of data to accurately model multitrillion ngram language spaces?
Is this part Nicks work?
Identifying the interim features a neural network learns on its way to an answer?
Are our brains using similar features?
Can we better understand our own minds by understanding the inner workings of chatgpt?
But how to evaluate more formal logic?
What is there was also a “semantic grammar” for language to complement the “syntactic grammar”?
What “model of the world” would a semantic grammar rely upon?
What does it take to produce “meaningful human language”?
What if you modulated temperature with each word?
Decreasing temperature when model is very confident and increasing when less?
how do you pull that out?
how do you enable talent to do the things they want to do?
Thinking about how do you pull that out?
How can we develop transformative tools for thought?
How to design the incentives?
small provocation in notes: Question: How do games manage positive and negative feedback?
How do they make negative feedback feel less discouraging and make you want to continue?
How did the computer animation industry get started; was it public research?
Seems like Pixar was able to build on their work?
If spaced repetition were different genres of video game, how would you play?
How do you structure a day?
A week?
How to solve the public goods problem?
when do you do this though?
How to design the platform for rapid experimentation with mnemonic mediums?
Improvements over spaced repetition?
Question: How do games manage positive and negative feedback?
How do they make negative feedback feel less discouraging and make you want to continue?
Problem: Encode stories in the mnemonic medium?
How do cards interact with each other?
Network structure of knowledge?
What happens when you remove a card, are there crucial lynchpin cards, effective network structures?
“Why should I care about memory?
Can’t I just look stuff up on the internet?
how to development judgement for what is important?
What was it that initially drew you and caught your interest?
The work is really about exploration of an openended question: how can we develop tools that change and expand the range of thoughts human beings can think?
Is this actually true though?
When an expert takes a paper and parses it for 2 minutes silently, then says that’s a good paper, what are they looking at?
What are the cascades happening inside their head?
What happened to it?
What tasks do we want to delegate?
Do we want to automate writing and coding with chatGPT?
What do these skills do for us?
Perhaps progress can measured by the rate which new fields are created?
How to 10x the rate at which new fields are created?
What programs would dramatically increase the rate of production of fruitful new fields?
If this wasn’t true, why would that be?
How would a person from this perspective think about this problem?
What does this type of person do?
What evidence supports your belief in this opinion?
What are some of the open questions to learn more about?
What are the sub tasks to complete for the problem?
What’s interesting to me right now?
Where are some of your strengths?
What would your employer say are some of your weaknesses?
How might they suggest you improve?
What are some of the important questions to explore that may increase your confidence on the matter?
Any new insights or connections to this idea?
Why do you believe that?
What might this suggest?
Does this require you to evaluate any of your beliefs?
Where did this evidence come from?
Does this strengthen or weaken the claim?
How might you reduce risk or uncertainty in this situation?
What percentage would you place on this belief?
What would you need to know to make it higher?
Asked, what would happen if in ten years you were told to stop using the device?
A reflection/memory/creativity support system that in ten years, would be something people would be unwilling to put down?
What would it look like?
What happens if I ask you ten years from now to stop using your personal communicator?
One day wandered out of his house, stared at the starfilled sky, and had an epiphany: In the face of the awesome celestial epic, what was the point of being depressed?
What makes something disruptive?
How to spot an idea?
What makes a great design?
What is it you want to learn?
And who do you want to learn from?
What if you applied this to the self?
Are you gonna get sued for this ad?
How to see the forest vs. the trees?
Also how should we talk about the progress of anything really?
But does that actually mean that you, you are making less progress?
What if spaced repetition was safety net?
Could you create naturalistic practice of things that otherwise might not?
What is you have a 10 minute practice and none of it is retrieval?
Can the review take place in the context of the thing that you originally read?
Can we feel like we’re back in the article somehow?
action you’re taking on these things?
what do you do with the front when you’re viewing the back?
Where should attention go?
And your eye look when you press the spacebar?
Just have this list of things that you’re supposed to do, what’s the theory of what you’re supposed to do to integrate the list?
Do we have anything powerful to say about that?
The writing group any insights from that experience that seem relevant to leverage here?
What are powerful ways of seeing?
How can we represent that behavior?
What if this is major limiter in our ability to deal with complex systems of today?
We’re arriving at a world where everything will have computation embedded in it, and how do we think about such a world?
Why are we building it?
What will it do to us?
How do we do this kind of modeling?
What does authorship look like in this medium?
How do you represent y=x^n?
how to redesign mathematics that draws on all our physical and intellectual capabilities?
What if cards are more dynamic spaces?
How will you handle various flavors of incorrect responses?
Where do you go next?
How did DARPA design a system for rapid acquisition of expertise?
What makes someone more likely to develop expertise?
Rank your questions: which is most critical?
If Ernest Hemingway, James Mitchener, Neil Simon, Frank Lloyd Wright, and Pablo Picasso could not get it right the first time, what makes you think that you will?
What if the major change in someone’s life is external influencers?
What role does internal mental state play?
What is the overlap and differences?
How to balance sociality and private?
May want to remember some things from a trip in context?
How so?
Or acknowledging the existence of human experience beyond the scope of its ideas?
How might one reconceptualize the primitives of a memory system to make prompt refinement a primary action?
Relatedly: could there be some smooth gestural interaction to indicate your emotional connection to the prompt, how much it seems to matter to you?
Then the system asks: did you remember that, or forget it?
What verbs (and schedule dynamics and feedback loops) are appropriate for nonmemory prompts like these?
How should you steer that set over time?
LLMs for prompt text: How might LLMs be used to make review more effective or stimulating or meaningful?
Maybe it would be interesting to have a little conversation about a prompt, as an exercise?
How might we help people internalize this kind of knowledge?
What interventions affect this behavior?
Could you situate a memory system directly in a context of use somehow?
Is there some sociallyembedded practice that might makes sense here?
What are metaphors one might want to use to interact with information over time?
How do we connect memory systems closer to the environments that we think and work?
How would tove jansson (propn) use (v) spaced repetition (n)?
Maybe it would be interesting to have a little conversation about a prompt, as an exercise?
What does this person mean to you?
why do you like this person?
how does it feel emotionally?
how much does it matter to you?
how can we make modifying, deleting, remixing prompts later as a primary primitive or action of memory systems?
What is the refresh rate?
First Summary: How can we better understand the effects of improved memory on a person’s life?
What if we focused on memory system designs where the primary intention is something else like supporting creativity, reasoning, insight and hypothesis generation, personal transformation?
What I structured all actions in this format from now on?
If the goal was to support creativity, reasoning, insight generation, hypothesis generation and personal transformation, how would you structure a memory space?
What interactions would you have with the ideas and concepts?
What are we really trying to get out of these systems?
can we establish a function which describes the dynamics of retrieval on memory?
What type of cognitive activities trigger different parts of the mind?
what would we really like to know, and how might we collect the data to know it?
how does the memory generalize to openended tasks in the subject?
how well can you explain these topics?
can you discover novel insights that connect to these ideas?
how do we improve transfer learning?
What if you tried iteratively summarizing the main point of a writer’s work by paragraph?
Then summarize those, eventually arriving at an acronym for the whole essay?
That is, if you study chapter 1 via a mnemonic text, can you learn chapter 2 more rapidly?
Accurately?
Deeply?
Creativity?
If we designed a new “memory system” with the sole aim of downstream impact on creative work, what would it look like?
If “extended contact” really is the primary goal, what fundamental “nouns” and “verbs” should we build a communications system around?
To what extent do people who “start to think of themselves as ‘a person who studies quantum computing’” then take action that reaffirms this identity outside of Quantum Country (work, study, meetup groups, twitter convos, etc)?
What about creativity?
Where do ideas come from?
So: can memory systems make us more insightful?
Presumably some kinds of prompts help here more than others—what characterizes that?
Are special synthesisoriented prompts helpful, or is the impact more a function of solidly understanding the basics?
If we designed a new “memory system” with the sole aim of downstream impact on creative work, what would it look like?
Would it involve retrieval practice at all?
First Action(s): How do we make long term goals feel present in the short term?
How certain are you that you want to do this?
How certain are you that you can do this?
How certain are you that you will do this?
Exploring the various mediums?
What are the activities that are divergent vs. convergent?
The actions you can take?
So perhaps the question instead is what are the divergence activities?
How much has been connected or added to some container?
Can you see salience of an idea change over time?
Maybe overexposure to a stimulus leading to desensitization?
Can you compress the entire experience into a single fundamental action?
At least the main chunk of the flow?
Also, what if you actually blocked the user from constantly interacting with the app?
After you’ve done the narrative experience in the morning, you can’t do another until night time?
ex: is this still important to you?
Why has spaced repetition as a daily practice for myself been hard?
And what might this suggest about the challenges with it becoming a widespread practice?
How do you want to structure the Commons thing?
What are you hoping to get out of it?
What are you hoping to learn?
what is knowledge tracing?
what makes a question transformative?
How can we surface the right questions at the right time to people to promote positive growth and reflection?
When you’re in a situation, what will you use?
How malleable are these dispositions?
The critical thinking?
Experience?
how much familiarity is needed to approach the question?
How does performance on critical thinking tests transpose into performance in other activities?
How do you perform in a more naturalistic sense?
They go home and read a newspaper, do they actually behave differently?
Very unforgiving, you break a two week streak, what’s the point of studying anymore?
Can you keep the motivation pure?
Not create some proxy motivator for learning?
But they are compelling because they give us a very simple decision function: how do I know when I should do the thing or whether I’m doing the thing enough?
So what if there was this aspect in scheduling where the activities are assigned a level of time or energy?
What if you take partial habits and wrap them all inside of a daily habit?
Push and Pull, how do these things interplay?
How to push this further this week?
Maybe a spaced repetition card you know what?
How to balance push and pull?
How can you make a practice that not only helps you remember but helps you deepen your understanding over time?
what are you wanting to do with these prompts?
What do you find yourself wanting to see?
Why?
What else is not actually externalized very well?
Is it not ironic, I think, that I have sought isolation here in these remote mountains, only to become reliant on modern technology?
Turning to Sonam, I ask, ’Would you like me to teach you the basics of Thangka painting, dear traveler?
How is he faring now, and have our separate paths brought the spiritual impact we’d once hoped?
Is it time to seek him out, to join together in our mission once more?
I wonder what prompted my thoughts in this direction?
’How would I find him?
How could this note have found me here in my hermitage?
A thousand questions run through my mind how, indeed, did he leave the note for me?
Could his path have connected with mine without my knowing?
Overwhelmed by a mixture of disbelief and joy, I respond, ’Tenzin, is that really you?
’Tenzin, do you remember that cloudy night we secretly stargazed by the monastery, questioning the possibilities of other worlds?
’Did you hear about that young pilot from Earth?
What if I’m not good enough?
What if I never make it?
Why would they be here, in our small village?
I couldn’t help but wonder; could this be my chance?
How would I manage to juggle this with all my responsibilities?
But what of my family, my village?
Lost in thought, I ask, ’Gava, do you think there’s any truth to the legends of the Earth Sanctuary?
Often mentioned?
Frequently mentioned?
what are the jobs to be done that these tools support?
What other next steps or moves seem worth exploring next to you?
However, perhaps it’s worth another look?
Is this the orange one?
Wait, has that always been there?
And so like, you’re sort of making all this progress and like maybe you drop something and it comes back in later, right?
Um, and so, uh, I, I think we can help with some of the stuckness by, um, not being dogmatic about sticking with any one of these things, but at least noticing where, um, when that’s happening and saying like, do we need to, um, go wide again?
[00:09:16] Or actually can we, can we stay with, uh, one particular thing?
Is there any other like project level material that you wanna discuss?
Um, You, you often, and I think this is good, like pick your head up and look around and see like, well, if I generalize this, like what’s the kind of space this could cover?
Is that if, if we’re imagining, like trying to imagine in detail, how will our proposed system help solve a person’s problem?
[00:14:59] In the case of the interviews you were trying to extract, like what are the problems people have, what are the strategies people have now?
How do they think about this?
I think you were trying to figure out what should I do next?
Uh, what is seeming most promising?
Uh, where do I feel like I have a path forward?
Where do, where am I feeling most excited?
One of the things you’ve been doing to your point is very much like what you’ve been doing with the research log, which is like you’re just trying to figure out like day to day, week to week, like, what should I be doing?
How would you categorize that?
Like in like, like for example, one of them was like, okay, what, what are some of the moves that we’ve taken?
[00:19:01] Matt: or intuitive level, um, what it is like to do these kinds of things?
[00:21:13] Um, what is it that people said?
Let him pass?
Um, do you have a sense of which of those three feels most alive to you?
[00:22:35] When I say 3d, do you, do you know what I’m referred to?
What, what was the word you used?
Did you say it was least clear?
Which one is like jumping out to me?
What?
Like that?
What is that?
Um, now the question is, do we have anything interesting to say about it?
Uh, uh, do we understand it clearly?
Can we articulate it clearly?
Um, so one way to frame this whole problem, I guess a very like, designerly way to frame this is like, uh, how might we apply, uh, large language models to the problem of figuring out what to do next?
Maybe that’s it, but then like what specific problem does that cause like when you’re trying to do some task, right?
Um, a way that we get around this is by kind of cheating and switching the distillation to like a greedy, almost streaming style operation where, um, you know, he suggests that we find threeish anchors that feel especially salient and then we like fit a line, right?
Um, and then you hold that line in, uh, the, the, your fit in mind, uh, and you keep reading and, uh, now you have only like one chunk that you’re holding, so you’re only consuming, you know, one of your end slots, right?
Right?
And then you’re kind of, you’re kind of developing your like hierarchy that way, right?
And then just trying to understand like, why are these related?
What, what are these trying to tell me?
But like a bunch of that is my reaction, you know?
Um, and then there’s some theme stuff, right?
It’s what have I really figured out?
[00:51:05] Uh, and then what next?
Where are we again?
And again, I’m asking where are we?
Have you seen this exhibit before, by the way?
So, but like, why is that bad?
[00:57:30] Andy: What was the obvious thing that you wanted to do following writing that note?
Um, do you want to do that or you want to keep pushing on this one?
Um, especially if like, I dunno, depending on the algorithms that you use, right?
Why am I feeling not compelled by that?
Can you, like, how does this thing relate to this thing?
Um, I’m reading over this thing, what am I trying to get out of it?
Or I’m doing this thing, what am I trying to get out of it?
Are you, what’s, what’s going on?
What’s going on?
[01:09:05] you know, what, what was actually wrong with what I did in this example that, that I showed you?
Like, what are we saying here?
Are we saying that like, well, I could have gotten this done faster, more easily, that I could have gotten a better answer?
[01:09:23] What are, what are we imagining?
Stopping myself to ask like, um, what problem am I really solving here?
Like, what am I trying to get to?
What would an answer look like?
What is it that I don’t like about this answer?
You pull, you pull on that string, um, and you ask like, what problem does that solve?
[01:13:27] Exactly?
How does it solve it?
Why does it solve it?
Like, why was it that previous solutions were insufficient?
[01:13:44] Matt: become even sharper?
What’s up?
What was that one thing?
So what, what problem is this solving for you?
Is it like letting you move through the conversation effectively?
I’m, I’m kind of wondering like, something about this felt powerful and like, what was it?
Right?
Um, where was that?
[01:27:52] What does that mean?
Like, what’s wrong with the themes that seem bad?
Um, mostly because, um, I think we haven’t pinned these things, right?
[01:33:47] If we’ve been saying, well, okay, what if we’re solving this other problem?
Is that okay?
Would you wanna meet here still?
Possible exploration: What if we constrained the agent to only observing actions from one main application (ex: google docs) and focused on observing the actions an expert makes in the process of providing feedback to someone else?
What if the novice acts as both teacher and learner?
Right?
And like, what is that they’re like, how has that like?
Had like this cascading set of consequences over time?
And so yeah, how do we help people construct, narrative or change the narrative frames that they’re coming from?
Over time?
It’s like a second ocean, motion ocean like?
So all of these little what, what am I interested in, right?
Do you remember what you did?
Like, do you remember what you did?
Like do you remember the the solutions the tools that you used in order to like, work through that time last time?
And so yeah, if I think this is kind of similar, because what you’re doing is you’re taking all of these concepts and like, where’s I’m feeling this I’m, oh, I’m starting to, I don’t know like maybe maybe when you’re depressed right?
So, okay, why don’t we hop you back?
Go back to that episode, and have you kind of like, replay some of the moments in that in that episode?
So what do you mean you say Semantic Web?
I’m wondering if there’s anything else so like, zoom back out?
But you can do so Okay, so remove the personality but what is good about the artificial agents in her?
Anything important from that example?
Trying to think if there’s anything else so the main question is, can you generate questions by trying to generate questions of similar quality to the?
The textbook or is it okay if it generates a different quality or flavor of question?
How will you handle various flavors of incorrect responses?
Where do you go next?
e.g., ’Why do you say that?
e.g., ’Is this always the case?
e.g., ’Why do you say that?
e.g., ’What is the counterargument?
e.g., ’But if…happened, what else would result?
e.g., ’Why do you think that I asked that question?
’, ’Why was that question important?
Revealing the issue: ’What evidence supports this idea?
Conceiving reasonable alternatives: ’What might be another explanation or viewpoint of the situation?
Evaluate those consequences: ’What’s the effect of thinking or believing this?
If “extended contact” really is the primary goal, what fundamental “nouns” and “verbs” should we build a communications system around?
Where are the powerful ideas about how people learn, feel, and act?
If we designed a new “memory system” with the sole aim of downstream impact on creative work, what would it look like?
What about creativity?
Where do ideas come from?
So: can memory systems make us more insightful?
Presumably some kinds of prompts help here more than others—what characterizes that?
Are special synthes is oriented prompts helpful, or is the impact more a function of solidly understanding the basics?
If we designed a new “memory system” with the sole aim of downstream impact on creative work, what would it look like?
Would it involve retrieval practice at all?
What if it doesn’t include retrieval practice?
What if we relied on some degree of computer supported retrieval to remind us of useful connections and to help us navigate a growing cluster of ideas?
I think one of the essential questions here is how do you create a world?
How do you help someone take a concept they believe is important and flesh out that concept in increasingly rich ways?
Generally you can apply noun verb constructs onto a card in order to apply mental operations on it if “extended contact” really is the primary goal, what fundamental “nouns” and “verbs” should we build a communications system around?
How we can learn to assimilate new representations, concepts, processes and techniques into our capability infrastructure?
How do we usually go about this?
What if we used ideas inspired by spaced repetition?
What is the goal of guided attention?
What is the equivalent in a like steampunk world?
What would be interesting to talk about?
What parts feel most important?
Workspaces have some way to expand and grow if I need more space?
Would that look like an unfolding?