Darmani

Replying toDeep learning as program synthesis

Darmani21d

Deep learning as program synthesis

Thank you! This really made your thesis click for me.

Replying toDeep learning as program synthesis

Darmani23d

Deep learning as program synthesis

Thanks for the clarification Zach!

Since I've been accused of lazy reading, I want to finish off the terminological discussion and put you in my shoes a bit. I get your motivation for preferring to call it "program synthesis" over "program induction," but it turns out that's an established term with about 60 years of history. Basically, to understand how I read it: replace every use of "program synthesis" with "techniques for searching constrained spaces of programs using things like SAT solvers and Monte Carlo." If you also replace "Daniel Selsam" with "researcher in SAT solvers who started off in a lab that uses Monte Carlo to generate assembly programs," then I actually... (read more)

-2

Replying toDeep learning as program synthesis

Darmani23d*

Deep learning as program synthesis

Hmmm. Okay then, I'd like to understand your point.

But first, can we clear up some terminological confusion?

From this comment, it seems you are using "program synthesis" in ways which are precisely the opposite of its usual meaning. This means I need to do substantial reverse-engineering of what you mean in every line, in addition to trying to understand your points directly.

These are narrow hypothesis classes where you, the practitioner, chose a representation embedding your assumptions about problem structure.

This is a very confusing thing to write. I think you're saying that various techniques are differentiated from program synthesis because they "choose a representation embedding assumptions about the problem structure." But can you point... (read more)

-3

Replying toDeep learning as program synthesis

Darmani23d*

Deep learning as program synthesis

Hi! I did my Ph. D. in a program synthesis lab that later become a mixed program synthesis / machine learning lab. "Machine learning is program synthesis under a different name" my advisor would say.

But my experienced turned out to not be very relevant to reading this post, because, I must say...I did not get what the point of this post is or the intended takeaways.

In genetic programming, there's a saying "Learning programs is generalized curve-fitting." And, indeed, the first chapter of "Foundations of Genetic Programming" is about trying to evolve a small AST that fits a curve. I gave a similar problem to my students as their intro to enumerative program... (read more)

-17

Replying toContra Dance as a Model For Post-AI Culture

Darmani1mo

Contra Dance as a Model For Post-AI Culture

I do not know hour they're handling it! Most of my demoscene knowledge comes from a single event by the CMU Computer Club in 2009.

Replying toContra Dance as a Model For Post-AI Culture

Darmani1mo

Contra Dance as a Model For Post-AI Culture

nit: When demosceners hand-write assembly, it's not because they're foregoing a compiler for the sake of it. They operate under performance and size constraints that require having assembly that compilers cannot output. They also often build for older platforms, with only comparatively primitive compilers available.

Replying toThe Doomers Were Right

Darmani4mo

The Doomers Were Right

Consensus view is that they were shielded by those who did invest in it.

I've written more about Y2K at https://www.lesswrong.com/posts/zvQdgfFEDFQQhDDuS/y2k-successful-practice-for-ai-alignment

Replying toThe Doomers Were Right

Darmani4mo

The Doomers Were Right

Small historical nit: China was actually ruled by the Mongols for most of the 1300s.

Berkeley Petrov Day

Darmani

5mo

In 1983, Stanislav Petrov did something brave and defiant. He didn't destroy the world.

Come celebrate his contribution as we find within ourselves the courage to continue to not destroy the world.

We'll be running the full ceremony from petrovday.com , complete with candles and big red buttons.

Location:
2307 Haste St

Berkeley, CA 94704

Replying toThe Gentle Romance

Darmani1y

The Gentle Romance

Am I the only one who finds parts of the early story rather dystopian? He sounds like a puppet being pulled around by the AI, gradually losing his ability to have his own thoughts and conversations. (That part's not written, but it's the inevitable result of asking the AI every time he encounters struggle.)

Replying toIf all trade is voluntary, then what is "exploitation?"

Darmani1y

If all trade is voluntary, then what is "exploitation?"

I made the grave error of framing this post in a way that invites a definition debate. While we are well familiar that a definition debate is similar to a debate over natural categories, which is a perfectly fine discussion to have, the discussion here has suffered because several people came in with competing categories.

I strongly endorse Ben's post, and will edit the top post to incorporate it.

If all trade is voluntary, then what is "exploitation?"

Darmani

Capitalism is a force that has lifted billions out of poverty, where even poor remote villagers enjoy luxuries that would have been unimaginable to medieval kings. When someone takes a job, even the worst job, it’s because both parties expect mutual gain. And yet companies routinely get accused of exploiting their workers for offering low pay and bad conditions, even if the pay and conditions are far better than the other available jobs. This sometimes results in protectionist laws that prevent those businesses from existing in the first place, making everyone worse off.

Given this, is there any meaningful concept that could be called “exploitation?”

I think there is.

In fact, I claim, it routinely... (read 1586 more words →)

Cambridge LW Meetup: When Science Isn't Enough

Tony Wang

Tony Wang, Darmani

This month’s LessWrong meetup will be on when science is and isn't enough. Discussion topics include:

What is science?
Where/why science has triumphed?
Where/why science falls short of optimality?
Pitfalls in communicating scientific knowledge.
What is the relationship between science and engineering?

Here is the reading list for this meetup (possibly to be expanded at the actual meetup). Time during the meetup will be dedicated to reading these pieces, so there’s no need to read them ahead of time.

Faster Than Science [5 mins]
A Failure, But Not Of Prediction [15 mins]

There will be some optional readings that you can engage in if you would like to read the articles ahead of time though.

As always, the meetup is open to... (read more)

AI Capabilities vs. AI Products

Darmani

It's a great time to start an AI company. You can create a product in days that would have been unbelievable just two years ago. Thanks to AI, there will soon be a one-person billion-dollar company.

It's a terrible time to start an AI company. The singularity is fast approaching, and any talent that enters the field should instead go work in AI alignment or policy. Living a monastic lifestyle is better than being dead.

But wait! My plan is to just build software that takes helps people count calories by taking pictures of their meals. How does counting calories help bring about the end of the world?

In the past, any company worthy of... (read 665 more words →)

Cambridge LW: Rationality Practice: The Map is Not the Territory

Darmani

Join us for a thought-provoking Less Wrong meetup on the topic of "The Map is not the Territory". As human beings, we rely heavily on mental models, concepts, and language to understand the world around us. However, we are often tempted into over-relying on them and failing to notice when we are reasoning in a way that does not correspond to physical realities.

During this meetup, we'll explore examples of ways in which we are tempted to think in terms of the map rather than the territory and how this leads us astray. We'll delve into the limitations of our maps and discuss how we can improve our understanding of the territory.

Whether you're... (read more)

Cambridge LW Rationality Practice: Being Specific

Tony Wang

Tony Wang, Darmani

Join us for our monthly rationality meetup, where we will work on the skill of being specific! Specificity is the art of reducing abstractions to better understand or communicate an idea.

In this meetup, we will practice being specific, and also discuss where in our lives we would benefit from being more specific, whether it's in our relationships, our work, our personal growth, or our efforts to solve big problems like climate change or AI alignment.

To prepare for the meetup, we invite you to read two posts from LessWrong:

"SotW: Be Specific" (https://www.lesswrong.com/posts/NgtYDP3ZtLJaM248W/sotw-be-specific)
"The Power to Solve Climate Change" (https://www.lesswrong.com/posts/rm2MueK3LxpRywE8N/the-power-to-solve-climate-change)

These posts offer some great insights into the importance and practicality of specificity, and will give... (read more)

Rationality Practice: Self-Deception

Darmani

Join us for another dive into rationality practice as we search for ways we may be deceiving ourselves and how to combat them. We will brainstorm techniques and strategies for spotting and correcting biased thinking, as well as ways to improve our mental habits and decision-making processes. Whether you're looking to improve your critical thinking skills or simply want to learn more about how to avoid falling prey to self-deception, this meetup is for you

This meetup will be based on the classic "How to Actually Change Your Mind" sequence, https://www.lesswrong.com/rationality#coGq3LC5Yn4vZiu6k . Read any two posts from the sequence before coming.

If you have trouble picking, try these:

https://www.lesswrong.com/s/5bZZZJ5psXrrD5BGb/p/43PTNr4ZMaezyAJ5o
https://www.lesswrong.com/s/5bZZZJ5psXrrD5BGb/p/CahCppKy9HuXe3j2i

Meetup description written with the help of ChatGPT. Cover image very much not created by AI.

Cambridge LW Meetup: Lifehacks

Tony Wang

Tony Wang, Darmani

Would you like to spend less time on your phone, get better sleep, work more efficiently, or build deeper connections with people? Do you also like to move fast and break things? If so, then this is the meetup for you.

For the final Cambridge LW meetup of 2022, we will be discussing lifehacks -- quick-and-dirty solutions to life's problems. Please join us to share your favorite tips/tricks for living better, and why you think they work. We will also have a meta-level discussion about the drawbacks of lifehacks, and discuss how the concept of "technical-debt" applies to each of our lives.

An example of a lifehack would be using GPT-3 to help write this announcement and DALL-E 2 to generate a nice picture to go along with it.

Snacks will be provided and there will be an informal dinner afterwards for those who wish to stick around.

Cambridge LW Meetup: Free Speech

Darmani

Free speech used to be seen as a universal good, but recently public opinion has started to shift on free speech, seeing it as a smokescreen for justifying harmful behavior. Ethics aside, it has now become common in the US to denounce people for voicing unpopular views, and society has changed for it.

Come in for a rousing discussion about the philosophy and practice of free speech today. There will be a presentation dispelling many myths about the legal aspects of free speech, followed by a discussion of free speech as a societal value.

Snacks will be provided.

Venue: In 100 Landsdowne, in the room behind the concierge. Concierge should also be able to direct you.

How much does cybersecurity reduce AI risk?

Darmani

AI is here, and AGI is coming. It's quite possible that any work being done now will be futile in comparison to reducing AI risk.

This is one of those things that's unsettling for me as someone who did a Ph. D. in a non-AI area of computer science.

But one of the main vectors by which a bootstrap AGI will gain power is by hacking into other systems. And that's something I can do something about.

Not many appreciate this, but unhackable systems are very possible. Security vulnerabilities occur when there is some broken assumption or coding mistake. They are not omnipresent: someone has to put them there. Software has in general gotten more... (read 175 more words →)

LESSWRONG
LW

LESSWRONG
LW

Leaky Delegation: You are not a Commodity

Y2K: Successful Practice for AI Alignment

Cause Awareness as a Factor against Cause Neutrality

If all trade is voluntary, then what is "exploitation?"

Darmani

Darmani

If all trade is voluntary, then what is "exploitation?"

AI Capabilities vs. AI Products

How much does cybersecurity reduce AI risk?

Nuclear Deterrence 101 (and why the US can't even hint at intervening in Ukraine)

What is a Counterfactual: An Elementary Introduction to the Causal Hierarchy

Y2K: Successful Practice for AI Alignment

Leaky Delegation: You are not a Commodity

Darmani

Leaky Delegation: You are not a Commodity

Y2K: Successful Practice for AI Alignment

Cause Awareness as a Factor against Cause Neutrality

If all trade is voluntary, then what is "exploitation?"

Darmani

Darmani

If all trade is voluntary, then what is "exploitation?"

AI Capabilities vs. AI Products

How much does cybersecurity reduce AI risk?

Nuclear Deterrence 101 (and why the US can't even hint at intervening in Ukraine)

What is a Counterfactual: An Elementary Introduction to the Causal Hierarchy

Y2K: Successful Practice for AI Alignment

Leaky Delegation: You are not a Commodity