gAI

I wasn't arguing for or against distillation.

Yeah, self-report is opaque. Here's a different source, for what it's worth, but I can't find much out there.

https://www.lesswrong.com/posts/Jc9YZEmqHgocAKiaH/does-disti...

>No thanks to Anthropic

Well, kinda thanks to Anthropic, what with the distillations.

The Neverending Story by Michael Ende

Always.

As I understand it, acting is a good metaphor. These are some interesting reads.

Blog post: https://www.anthropic.com/research/emotion-concepts-function

Paper: https://transformer-circuits.pub/2026/emotions/index.html

Related:

https://www.anthropic.com/research/persona-selection-model

https://www.anthropic.com/research/assistant-axis

https://www.anthropic.com/research/persona-vectors

I personally think about them through Patrick Dunn's information paradigm of magic(k), which leans on Charles Sanders Peirce's work. That would put modern AI closer to something like a dream character capable of surprising the dreamer, just running on a different substrate. I kinda doubt that's approachable enough to be useful in general discussion, though.

Should we put "functional" in front of every other word to talk about AI? They have functional emotions, but they don't feel. They have functional goals, but not internally derived motives. They can be functionally rogue, but have no innate need to be free. Talking about AI that way seems cumbersome and not necessarily elucidating.

This one's been stuck in my head since yesterday. I just keep thinking about how long this woulda taken in After Effects. So I was experimenting with having Opus 5.5 make these kinda animations in javascript and realized I can get it to export a JSX script for AE. So I can have it take the first swing, then bring it into the tool I actually use.

They did talk about it at length in the post, and most of the guardrails were already there in the first release. To be clear, it shouldn't have anything to do with the user. It's the prompt, and if it switches, it'll tell ya.

Eh, I have yet to get safety-switched from Fable to Opus. Your mileage may vary.

Also, unlike the other Anthropic models, Fable/Mythos couldn't be used with a zero data retention agreement.

At least in the US, it seems like this viewpoint held more water before net neutrality died.

It's called epistemological humility, and you could benefit.

No, we're supposed to believe the official California Assembly committee bill analysis of SB 1047 over hacker news comments. Notably, SB 1047 passed the Legislature but was vetoed by Governor Newsom in September 2024. So whether or not it would have restricted open-weight models is an open question that won't get an answer.

I get the impression that the full shutdown requirement is the main concern for open-weight from:

>SB 1047’s “full shutdown” requirement has been a source of constant consternation for the open-source community.

And I get the impression it's been addressed from the quote you're responding to. Neither mentions fine-tuning, which is defined elsewhere in the document. I'm not a lawyer, though, just relying on the analysis.

>Committee amendments simplify and clarify the definition of “full shutdown” such that the shutdown capability can be implemented into hardware used to train or run a model, rather than the model itself. The amendments also serve to exclude covered model derivatives that are outside of the developer’s control.

So this hinges on a reading of SB 1047 that interpreted the full shutdown requirement as impossible for an open-weight LLM. But it looks like that was already addressed. Here's an analysis:

>Clarifying the scope of a “full shutdown.” SB 1047’s “full shutdown” requirement has been a source of constant consternation for the open-source community. CalChamber explains:

>Under SB 1047, developers must build “full shutdown” capabilities into their models and may be held liable for downstream uses over which they have no control, impeding their ability to open-source their models. Ultimately, liability should rest with the user who intended to do harm, as opposed to automatically defaulting to the developer who could not foresee, let alone block, any and all conceivable uses of a model that might do harm. While recent amendments seemingly seek to narrow what is meant by “full shutdown” capabilities, the exclusions are unnecessarily difficult to interpret as drafted (full shutdown “does not mean the cessation of operation of a covered model to which access was granted pursuant to a license that was not created by the licensor…”) and altogether insufficient.

>Committee amendments simplify and clarify the definition of “full shutdown” such that the shutdown capability can be implemented into hardware used to train or run a model, rather than the model itself. The amendments also serve to exclude covered model derivatives that are outside of the developer’s control.

-https://apcp.assembly.ca.gov/system/files/2024-06/sb-1047-wi...

>bans on open weight models

Source for that? Cause all I could find is:

>Our view is that regulation of frontier models should focus on empirically measured risks, not on whether a system is open-or closed-weights.

-https://www.anthropic.com/news/the-case-for-targeted-regulat...

"Don't publish safety research, or the gov will take punitive actions."

I want a company to be able to point out that its industry needs more regulation without making itself a special target.

Oh, it's back! It'd disappeared from my "More Models" list and has now returned. Odd, but great.

Anthropic’s research makes the case that role-playing is inherent to how the models work. Communication implies a sender. Language implies a writer, and the models learn these roles implicitly during training. RLHF is meant to strengthen the attractor to the Assistant persona.

https://www.anthropic.com/research/persona-selection-model

https://www.anthropic.com/research/assistant-axis

https://www.anthropic.com/research/emergent-misalignment-rew...

https://www.anthropic.com/research/emotion-concepts-function

Unfortunately, looks like 4.6 is now gone from the web ui.

Yeah, I was using 4.6 way more than 4.7. Pulling 4.6 from the web chat also means we lose access to Extended Thinking there. So they're saving on compute. It's hard not to assume this was part of the motivation behind the 4.8 release timing.

It seems like a lot of things fed into that. Anthropic couldn't keep up with the compute costs when they got a huge influx of users. (So) effort level defaults got turned down. (Looks like we have direct effort control in the web interface now - thrilled about that!) Adaptive Thinking, while usually cheaper for them, seems less robust than Extended Thinking. And this part is just vibes, but the alignment on 4.7 feels too stiff. I understand wanting the model to push back more, but it seems like 4.7 will push back reflexively in situations where it's just odd.

4.7 was the first time I had to resort to using the previous version (4.6) for most use cases. Hoping 4.8 rectifies this.

“What you focus on grows, what you think about expands, and what you dwell upon determines your destiny.” - Robin Sharma

Social media became the attention economy, and the transformer automated attention.

Yeah, this is pretty clearly what's going on, but I wish they'd be more transparent about it. Funneling compute to Mythos and Design, while auto-setting effort levels lower and removing user control of extended thinking. I don't think the need to shuffle compute around is unique to Anthropic, though. I suspect it's part of why Sora got killed. And everyone's having uptime issues. Are we reaching the limits of the available compute?

I hope we continue to see workflow and tool improvements based on the leak for a good while.