Typo:
We sequencing a typical sample to between one and two billion reads.
Should maybe be “We will be sequencing...”?
DanielFilan
Karma: 8,480
AXRP Episode 38.3 - Erik Jenner on Learned Look-Ahead
Wait I’m a moron and the thing I checked was actually whether it was an exponential function, sorry.
Votes cost quadratic points – a vote strength of “1” costs 1 point. A vote of strength 4 costs 10 points. A vote of strength 9 costs 45.
FYI this is not a quadratic function.
Dojo Organizations What organizations are you aware of that are providing some kind of rationality dojo format (courses focused on improving the skill of rationality)?
Seems like the stuff after “Dojo Organizations” should be on a new line.
About how often do you use LLMs like ChatGPT while active?
What does “while active” mean in this question?
AXRP Episode 39 - Evan Hubinger on Model Organisms of Misalignment
If one wants to investigate [the Alignment of Complex Systems research group] further, he has an AXRP podcast episode, which I haven’t listened to.
Note that if you want to investigate further but would rather read a transcript than watch a video, AXRP has you covered.
AXRP Episode 38.2 - Jesse Hoogland on Singular Learning Theory
Yeah but a bunch of people might actually answer how their neigbours will vote, given that that’s what the pollster asked—and if the question is phrased as the post assumes, that’s going to be a massive issue.
AXRP Episode 38.1 - Alan Chan on Agent Infrastructure
So I guess 1.5% of Americans have worse judgment than I expected (by my lights, as someone who thinks that Trump is really bad). Those 1.5% were incredibly important for the outcome of the election and for the future of the country, but they are only 1.5% of the population.
Nitpick: they are 1.5% of the voting population, making them around 0.7% of the US population.
If you ask people who they’re voting for, 50% will say they’re voting for Harris. But if you ask them who most of their neighbors are voting for, only 25% will say Harris and 75% will say Trump!
Note this issue could be fixed if you instead ask people who the neighbour immediately to the right of their house/apartment will vote for, which I think is compatible with what we know about this poll. That said, the critique of “do people actually know” stands.
she should have picked Josh Shapiro as her running mate
Note that this news story makes allegations that, if true, make it sound like the decision was partly Shapiro’s:
Following Harris’s interview with Pennsylvania Governor Josh Shapiro, there was a sense among Shapiro’s team that the meeting did not go as well as it could have, sources familiar with the matter tell ABC News.
Later Sunday, after the interview, Shapiro placed a phone call to Harris’ team, indicating he had reservations about leaving his job as governor, sources said.
Oh except: I did not necessarily mean to claim that any of the things I mentioned were missing from the alignment research scene, or that they were present.
When I wrote that, I wasn’t thinking so much about evals / model organisms as stuff like:
putting a bunch of agents in a simulated world and seeing how they interact
weak-to-strong / easy-to-hard generalization
basically stuff along the lines of “when you put agents in X situation, they tend to do Y thing”, rather than trying to understand latent causes / capabilities
AXRP Episode 38.0 - Zhijing Jin on LLMs, Causality, and Multi-Agent Systems
Yeah, that seems right to me.
A theory of how alignment research should work
(cross-posted from danielfilan.com)
Epistemic status:
I listened to the Dwarkesh episode with Gwern and started attempting to think about life, the universe, and everything
less than an hour of thought has gone into this post
that said, it comes from a background of me thinking for a while about how the field of AI alignment should relate to agent foundations research
Maybe obvious to everyone but me, or totally wrong (this doesn’t really grapple with the challenges of working in a domain where an intelligent being might be working against you), but:
we currently don’t know how to make super-smart computers that do our will
this is not just a problem of having a design that is not feasible to implement: we do not even have a sense of what the design would be
I’m trying to somewhat abstract over intent alignment vs control approaches, but am mostly thinking about intent alignment
I have not thought that much about societal/systemic risks very much, and this post doesn’t really address them.
ideally we would figure out how to do this
the closest traction that we have: deep learning seems to work well in practice, altho our theoretical knowledge of why it works so well or how capabilities are implemented is lagging
how should we proceed? Well:
thinking about theory alone has not been practical
probably we need to look at things that exhibit alignment-related phenomena and understand them, and that will help us develop the requisite theory
said things are probably neural networks
there are two ways we can look at neural networks: their behaviour, and their implementation.
looking at behaviour is conceptually straightforward, and valuable, and being done
looking at their implementation is less obvious
what we need is tooling that lets us see relevant things about how neural networks are working
such tools (e.g. SAEs) are not impossible to create, but it is not obvious that their outputs tell us quantities that are actually of interest
in order to discipline the creation of such tools, we should demand that they help us understand models in ways that matter
see Stephen Casper’s engineer’s interpretability sequence, Jason Gross on compact proofs
once we get such tools, we should be trying to use them to understand alignment-relevant phenomena, to build up our theory of what we want out of alignment and how it might be implemented
this is also a thing that looking at the external behaviour of models in alignment-relevant contexts should be doing
so should we be just doing totally empirical things? No.
firstly, we need to be disciplined along the way by making sure that we are looking at settings that are in fact relevant to the alignment problem, when we do our behavioural analysis and benchmark our interpretability tools. This involves having a model of what situations are in fact alignment-relevant, what problems we will face as models get smarter, etc
secondly, once we have the building blocks for theory, ideally we will put them together and make some actual theorems like “in such-and-such situations models will never become deceptive” (where ‘deceptive’ has been satisfactorily operationalized in a way that suffices to derive good outcomes from no deception and relatively benign humans)
I’m imagining the above as being analogous to an imagined history of statistical mechanics (people who know this history or who have read “inventing temperature” should let me know if I’m totally wrong about it):
first we have steam engines etc
then we figure out that ‘temperature’ and ‘entropy’ are relevant things to track for making the engines run
then we relate temperature, entropy, and pressure
then we get a good theory of thermodynamics
then we develop statistical mechanics
exceptions to “theory without empiricism doesn’t work”:
thinking about deceptive mesa-optimization
CIRL analysis
lesson of above: theory does seem to help us analyze some issues and raise possibilities
Would being in a room with people who are vaping have the same benefits as the fog machine? Obviously it has downsides of smell and other additives, but still—I think this should predict that people maybe don’t get airborne illnesses at vaping conventions.