This post is a not a so secret analogy for the AI Alignment problem. Via a fictional dialog, Eliezer explores and counters common questions to the Rocket Alignment Problem as approached by the Mathematics of Intentional Rocketry Institute.
MIRI researchers will tell you they're worried that "right now, nobody can tell you how to point your rocket’s nose such that it goes to the moon, nor indeed any prespecified celestial destination."
The history of science has tons of examples of the same thing being discovered multiple time independently; wikipedia has a whole list of examples here. If your goal in studying the history of science is to extract the predictable/overdetermined component of humanity's trajectory, then it makes sense to focus on such examples.
But if your goal is to achieve high counterfactual impact in your own research, then you should probably draw inspiration from the opposite: "singular" discoveries, i.e. discoveries which nobody else was anywhere close to figuring out. After all, if someone else would have figured it out shortly after anyways, then the discovery probably wasn't very counterfactually impactful.
Alas, nobody seems to have made a list of highly counterfactual scientific discoveries, to complement wikipedia's list of multiple discoveries.
To...
Did I just say SLT is the Newtonian gravity of deep learning? Hubris of the highest order!
But also yes... I think I am saying that
Text of post based on our blog post as a linkpost for the full paper which is considerably longer and more detailed.
Neural networks are trained on data, not programmed to follow rules. We understand the math of the trained network exactly – each neuron in a neural network performs simple arithmetic – but we don't understand why those mathematical operations result in the behaviors we see. This makes it hard to diagnose failure modes, hard to know how to fix them, and hard to certify that a model is truly safe.
Luckily for those of us trying to understand artificial neural networks, we can simultaneously record the activation of every neuron in the network, intervene by silencing or stimulating them, and test the network's response to any possible...
It's a sparse autoencoder because part of the loss function is an L1 penalty encouraging sparsity in the hidden layer. Otherwise, it would indeed learn a simple identity map!
Hypothesis: The truth, defined as the information with 100% credibility calculated by the HITS algorithm in accordance with the consensus theory of truth, is an attractor in the information space.
The post Knowledge Base 2: The structure and the method of building describes the possibility of building a knowledge database using crowdsourcing. After adding initial knowledge using crowdsourcing, also artificial intelligence and robots could add knowledge to this database during the execution of tasks, for example, while finding answers to the questions of their users or users of the database. People, computer programs, and robots could collaborate with each other to add and evaluate knowledge, increasing the amount and credibility of useful knowledge. Thus, this database could be a common interface for the exchange...
Epistemic status: party trick
One famed feature of Bayesian inference is that it involves prior probability distributions. Given an exhaustive collection of mutually exclusive ways the world could be (hereafter called ‘hypotheses’), one starts with a sense of how likely the world is to be described by each hypothesis, in the absence of any contingent relevant evidence. One then combines this prior with a likelihood distribution, which for each hypothesis gives the probability that one would see any particular set of evidence, to get a posterior distribution of how likely each hypothesis is to be true given observed evidence. The prior and the likelihood seem pretty different: the prior is looking at the probability of the hypotheses in question, whereas the likelihood is looking at...
Using a discrete hypothesis space avoids big parts of the problem.
Only if there is a "natural" discretisation of the hypothesis space. It's fine for coin tosses and die rolls, but if the problem itself is continuous, different discretisations will give the same problems as different continuous parameterisations.
In general, when infinities naturally arise but cause problems, decreeing that everything must be finite does not solve those problems, and introduces problems of its own.
The difference between EU and US healthcare systems
This comment may be replied to by anyone.
Other comments are for the discussion group only.
A friend has spent the last three years hounding me about seed oils. Every time I thought I was safe, he’d wait a couple months and renew his attack:
“When are you going to write about seed oils?”
“Did you know that seed oils are why there’s so much {obesity, heart disease, diabetes, inflammation, cancer, dementia}?”
“Why did you write about {meth, the death penalty, consciousness, nukes, ethylene, abortion, AI, aliens, colonoscopies, Tunnel Man, Bourdieu, Assange} when you could have written about seed oils?”
“Isn’t it time to quit your silly navel-gazing and use your weird obsessive personality to make a dent in the world—by writing about seed oils?”
He’d often send screenshots of people reminding each other that Corn Oil is Murder and that it’s critical that we overturn our lives...
Raw spinach in particular also has high levels of oxalic acid, which can interfere with the absorption of other nutrients, and cause kidney stones when binding with calcium. Processing it by cooking can reduce its concentration and impact significantly without reducing other nutrients in the spinach as much.
Grinding and blending foods is itself processing. I don't know what impact it has on nutrition, but mechanically speaking, you can imagine digestion proceeding differently depending on how much of it has already been done.
You do need a certain amount of...
Hey Bogdan, I'd be interested in doing a project on this or at least putting together a proposal we can share to get funding.
I've been brainstorming new directions (with @Quintin Pope) this past week, and we think it would be good to use/develop some automated interpretability techniques we can then apply to a set of model interventions to see if there are techniques we can use to improve model interpretability (e.g. L1 regularization).
I saw the MAIA paper, too; I'd like to look into it some more.
Anyway, here's a related blurb I wrote:
...Project: Regularizati
This article is part of a series of ~10 posts comprising a 2024 State of the AI Regulatory Landscape Review, conducted by the Governance Recommendations Research Program at Convergence Analysis. Each post will cover a specific domain of AI governance (e.g. incident reporting, safety evals, model registries, etc.). We’ll provide an overview of existing regulations, focusing on the US, EU, and China as the leading governmental bodies currently developing AI legislation. Additionally, we’ll discuss the relevant context behind each domain and conduct a short analysis.
This series is intended to be a primer for policymakers, researchers, and individuals seeking to develop a high-level overview of the current AI governance space. We’ll publish individual posts on our website and release a comprehensive report at the end of this series.
For the first point, there's also the question of whether 'slightly superhuman' intelligences would actually fit any of our intuitions about ASI or not. There's a bit of an assumption in that we jump headfirst into recursive self-improvement at some point, but if that has diminishing returns, we happen to hit a plateau a bit over human, and it still has notable costs to train, host and run, the impact could still be limited to something not much unlike giving a random set of especially intelligent expert humans the specific powers of the AI system. Additio...