Albert Wenger

@albertwenger

99 Following

1174 Followers

Albert Wenger pfp

Philosophy Mondays: Values https://continuations.com/philosophy-mondays-values

3 replies

16 recasts

85 reactions

Albert Wenger pfp

Reasonable people can disagree on policies such as tariffs but abducting foreigners and threatening citizens with the same is straight up totalitarian.

1 reply

3 recasts

13 reactions

Albert Wenger pfp

Reasonable people can disagree on policies such as tariffs but abducting foreigners and threatening citizens with the same is straight up totalitarian.

1 reply

3 recasts

13 reactions

Albert Wenger pfp

Super excited about this potential path to "actually open" AI

1 reply

1 recast

6 reactions

Albert Wenger pfp

Blog post: Some Thoughts on the Trump Presidency https://continuations.com/some-thoughts-on-the-trump-presidency

7 replies

25 recasts

122 reactions

Albert Wenger pfp

Philosophy Mondays: Randomness, Replication, Selection and Free Will https://continuations.com/philosophy-mondays-randomness,-replication,-selection-and-free-will

0 reply

0 recast

1 reaction

Albert Wenger pfp

Wrote a post on Musk as a potential scapegoat for Trump: A Scenario https://continuations.com/a-scenario

1 reply

0 recast

5 reactions

Albert Wenger pfp

Yes I saw that also. It is yet more evidence that we cannot leave values until post training. Have to include them in pre-training.

0 reply

0 recast

1 reaction

Albert Wenger pfp

Philosophy Mondays: Judgment https://continuations.com/philosophy-mondays-judgment

2 replies

3 recasts

42 reactions

Albert Wenger pfp

I believe there is a universal humanist core but the process of coming up with that can't be emergent from models but must be human led.

1 reply

0 recast

2 reactions

Albert Wenger pfp

So about that ... this fascinating paper shows that models develop their own value systems and what emerges is, well, problematic arxiv.org/abs/2502.08640

0 reply

0 recast

0 reaction

Albert Wenger pfp

Remember when Sam Altman said that AI systems could just "learn the collective moral preferences" in response to a question by Jack Kornfield about values? https://youtu.be/hn1Y6GVWUV0?si=wLl9t023O8eYREXm&t=750

2 replies

0 recast

2 reactions

Albert Wenger pfp

Philosophy Mondays (yes I realize it's Tuesday today): Choice https://continuations.com/philosophy-mondays-choice

0 reply

1 recast

4 reactions

Albert Wenger pfp

A Humanist Vision for AI -- Speech at the Paris AI Summit (Alternate History) https://continuations.com/a-humanist-vision-for-ai

0 reply

0 recast

6 reactions

Albert Wenger pfp

"The right to be represented by a bot" needs to become a fundamental right for all humans. We need to be able to mediate our interactions with systems through AI that we trust. I also cover this idea in my book The World After Capital https://worldaftercapital.gitbook.io/worldaftercapital/part-four/informational#bots-for-all-of-us

1 reply

0 recast

3 reactions

Albert Wenger pfp

I gave a talk 10 years ago titled at BIG Idea and a Bot Idea about basic income and the right to be represented by an AI bot that has held up well https://www.youtube.com/watch?v=t8qo7pzH_NM&t=1s

0 reply

3 recasts

5 reactions

Albert Wenger pfp

Philosophy Mondays: Action https://continuations.com/philosophy-mondays-action

0 reply

0 recast

2 reactions

Albert Wenger pfp

Announcing the @usv investment in Miren https://blog.usv.com/miren

0 reply

2 recasts

19 reactions

Albert Wenger pfp

Philosophy Mondays: Human-AI Collaboration https://continuations.com/philosophy-mondays-human-ai-collaboration

1 reply

0 recast

4 reactions

Albert Wenger pfp

A simple proposal: Inner Alignment through Values-Based Pretraining https://continuations.com/inner-alignment-through-values-based-pretraining (includes an offer of grants for research)

0 reply

0 recast

1 reaction