List of questions I have about the world. More directed questions about topics can be found in various notebooks.
If this list doesn’t quench your thirst for questions, feel free to check out Gwern’s, Kanjun’s, Alexey’s, or Patrick Collison’s. If any of these pique your interest, start a conversation with me!
Questions
- How do you do good science with dumb questions? I don’t know how to word this, but it feels like a lot of the amazing science is based on questions that people just stopped questioning. Especially salient is the movement of the cosmos and our place in the universe which was a “solved” problem before science came along and knocked us off our pedestal.
- Is life continuous? Is it that quantum particles can jump between states? How is an electron the smallest measure of charge? Why are certain domains discrete and others continuous? Are “continuous” domains ones where we don’t have enough resolution?
- How can you measure tradeoff between conformity and diversity? There is strength in diversity, but it seems like there is also a lot of social push for conformity. Is there an optimal balance? What kind of metrics would someone use to measure these?
- Does evolution push for diversity and inclusion?
- Do humans suffer from silent data corruption? If you flip any one base pair, are we really that fucked though? Very different compute paradigm than LLMs currently
- Has human genome mutation rate dropped throughout history? When you run evolutionary algorithms that can adapt their mutation, organisms eventually converge to no mutation as that is the only stable state. In such a case, why haven’t we converged to that? Is it a controls problem where it is harder in the real world, is no mutation something bad enough to prevent it from happening?
- How did life arise from abiotic components? Assembly theory might be a key here in the complexification of elements.
- Why does nature restore focus? Could just staring up at the vastness of clouds also serve as a restoration mechanic?
- Has technology domesticated us? Not fully checked yet, but story about sheep brains becoming smaller after domestication, and how that might happen to humans? Do all animal brains become smaller after domestication? Might ours also after we rely on AI for everything? Is this a valid fear?
- Can LLMs be used as a form of networking service? Goodside has shown some use in finding interesting people to follow on Twitter, but can this idea be extended? Perhaps it can also help enrich reading lists too!
- How did taste evolve? Better yet, how in the world do we convert foods into taste?
- Why are there only 4 nucleic acids in DNA? How did these evolve and why do these building blocks work so well?
- Why do genetic algorithms use bit-strings? Are there more robust underlying representations we might be able to use? Could we evolve more powerful genetic operators? Kenneth Stanley’s last paper at OpenAI kind of explores these ideas in more detail.
- Can we “prompt engineer” humans? In the same way we now prompt AI models to elicit better responses, might the same concept be applicable to humans? Coud we iterate towards to infinity by letting models prompt themselves?
- Is it possible to train AI smarter than ourselves? Might we be limiting it by making it use our language (both linguistic and mathematical). Could AI ever develop something beyond our furthest comprehension?
- Are current AI models poisoning the datasets?
- Is effective altruism overfitting?
- Are we not giving children enough autonomy? Here are some sources.
- Are humans intrinsically altruistic? Is altruism positive?
- What would non-coercive education look like? Is this truly better than the conforming one we have today?
- Are beauty or ethics objective? If so, how close can we get?
- How close are we to simulating a universe?
- How do we escape the Great Stagnation?
- What impact does telling someone they have a disease have on lifespan? This is not ethical. Very not ethical. But, it would be interesting to figure out.
- Can the internet ever eliminate geographically regional advantages? Current thoughts are that it might replace it with its own form of “virtual niche” regional advantage? Link to paper where elite universities are losing their edge.
- When does correlation become causation? My statistics professor linked me to a textbook on Causal Inference, but I haven’t had time to look into it yet.1
- Why are curiosity and goal-orientation often at odds? I’ve been talking with Kanjun about this particular question, but I still haven’t found a satisfactory answer.
- How can we quantify diversity? Just how “good” of a thing is it? Currently exploring in Quantifying Diversity.
- Is political/ideological unification bad? Thinking of states as individuals in a population based meta-heuristic searches like genetic algorithms, would it not be best to have diverse set of beliefs? That said, too much fragmentation also seems bad because it leaves an inability to deeply explore any of the proposed paths.
- Are there placebos in psychology?
- Do “better” forms of note-taking result in better works? Is this something that can be generalized across fields? Zettelkastens and note-taking apps with cog logos are all the craze, but are their paradigms truly better or worth the time investment? For those exploring this topic, I highly recommend looking into Andy Matuschak’s writings.
- How much of an effect do indirect genes have? Parents have genes they didn’t pass on, but they express the behavior which could lead to the child having the phenotype without the genotype. Is this a common occurrence? How strong of an impact does it have?
- What are the differences between finite and infinite games? How do we encourage infiniteness? Explored with Chris Lakin.
- How do we solve the tragedy of the commons? Based on the original paper written by Garrett Hardin
- How can we increase individual/group creative capabilities?
- How does supplanting thinking affect knowledge? Recent rise of Codex, DALLE-2, and GPT-3 suggest a feature where it’s more important to have creative ideas rather than technical prowess. Do these free us to focus on higher level issues? [@barrBrainYourPocket2015]