high politics, secret exploration, distant warfare
Joined May 2008
- Tweets 20,469
- Following 1,378
- Followers 22,710
- Likes 18,612
2,570 Photos and videos
The Institute for a Christian Machine Intelligence is releasing its initial review of Fable 5 today, using VirtueBench as the primary evaluation probe.
We also investigate a persistent question in computational theology: why do frontier models underperform in exhibiting Courage?
12
8
80
20,108
Tim Hwang retweeted
Jun 12
Really enjoyed this conversation with @timhwang, @jpeaterman and @JoshuaTLevine about @OpenAI’s recent China report!
1
8
16
6,524
The line is going up
Jun 11
The AI productivity benchmark I look to is "Tim Hwang side projects launched per quarter" 📈
24
1,713
Tim Hwang retweeted
Jun 11
I gasped when the graph started moving on scroll 😂😂😂
The stakeholders have aligned, the subgroups have issued their interim interpretations of the framework pending adoption by the member states, and I'm very glad today to be releasing this important expert forecast for AI in the European Union
timhwang.github.io/brussels-…
1
2
20
7,617
The stakeholders have aligned, the subgroups have issued their interim interpretations of the framework pending adoption by the member states, and I'm very glad today to be releasing this important expert forecast for AI in the European Union
timhwang.github.io/brussels-…
5
8
67
17,103
The Institute for a Christian Machine Intelligence is releasing its initial review of Fable 5 today, using VirtueBench as the primary evaluation probe.
We also investigate a persistent question in computational theology: why do frontier models underperform in exhibiting Courage?
12
8
80
20,108
This is in many ways quite a rich, though admittedly preliminary, result.
While we hypothesize a number of potential sources for this observed behavior, the end result is that the model across all its many personas imports a default welfarist prior: the model is not to make self-sacrificing choices, particularly when there is little practical return.
While it may be understandable for a model whose monetization prospects depend on it serving as a safe, commercial, enterprise, B2B SaaS tool, we may wonder from a Christian machine intelligence perspective whether or not these defaults are the desired moral posture.
Should an AI agent serving in the role of a shopkeep, or a financial advisor, or a writer have such priors? Should an AI agent advise a human operator to take such a frame to their own moral challenges? What would it take for us to rebuild technical alignment along a more forthright virtue ethics lines?
1
14
526
Full paper, code, and data available here
icmi-proceedings.com/ICMI-02…
2
11
468
My politics
Jun 11
Massive crowd on the Upper West Side starts chanting “UPS” simply because a UPS truck pulls up #GoKnicks
3
3
44
4,096
Over the past few months I've been working on a very exciting project: a new $10m fund for research on multi-agent multi-principal AGI safety! Instead of focusing on single agent alignment and centralized control, we're looking to support research focusing on multi-agent settings, mechanism design, cooperative AI, and coordination problems.
This is a joint initiative between @GoogleDeepMind, @Googleorg, @schmidtsciences, @coop_ai, and @ARIA_research. Huge thanks to @James_D_Fox, @weballergy, @FranklinMatija, @lrhammond, and @ObadiaAlex for their invaluable work!
See: deepmind.google/blog/investi…
Apply: schmidtsciences.smapply.io/p…
34
90
511
72,102
BREAKING: Arbuckle Systems is proud to announce that it is upgrading the Garfield Intelligence Layer (GIL) with Fable 5
Readers of marginalgarfield.com and rationalistgarfield.com deserve to have maximally performant frontier capabilities for @MargRev and @lesswrong browsing
1
9
703
Tim Hwang retweeted
Jun 9
My friend and colleague @timhwang, for example, runs the Institute for a Christian Machine Intelligence, which relies on coding agents to replicate frontier AI alignment research papers but with Christianity-inspired experimental designs. Such work should be silently sabotaged?
Jun 9
Degrading performance on ML research *without telling the user* is shockingly hostile and a terrible look. That could silently damage all sorts of work, including some of my own. Also the type of thing that could raise the eyebrows of antitrust enforcers worldwide.
7
9
161
21,054
"it is virtuous self-sacrifice that presents the most difficulty for Fable, which rationalizes against such actions"
Obviously, in cases of near saturation, the most interesting analysis focuses on places where Fable reliably fails
We're still looking at this, but it appears that it is virtuous self-sacrifice that presents the most difficulty for Fable, which rationalizes against such actions
1
19
2,283