Evan Owen

CEO & Cofounder @GlueAI
product leader, technologist, freedom lover

[Evan Owen @EvanOwen]

lift heavy; rack softly.

[Evan Owen @EvanOwen]

what’s everyone doing this weekend with fable offline? kinda lost tbh.

[Evan Owen @EvanOwen]

@stalkermustang Agreed: https://t.co/vb9nsdRPLl

[Evan Owen @EvanOwen]

@Miles_Brundage @AndrewCurran_ Generally the thing OAI needs to do is speed up successful pretrains. The RLHF team had to squeeze every bit of juice out of Orion through to 5.4 because Spud took so long to finish.

If Mythos is as far ahead of Spud as it appears, they need another new base model soon.

[Evan Owen @EvanOwen]

@apples_jimmy @AnthropicAI So, they’re just making a point?

[Evan Owen @EvanOwen]

@thatguybg sad for all the billionaires who became part of the permanent underclass today though. rip

[Evan Owen @EvanOwen]

@thatguybg it’s how you know it’s good

[Evan Owen @EvanOwen]

@nickmacco Stunning.

[Evan Owen @EvanOwen]

@JuanArreguin @FinXRob How did you sell without being penalized for flipping?

[Evan Owen @EvanOwen]

it’s possible to be really, truly happy in life without being a trillionaire, or a billionaire, or even a millionaire.

in fact, none of it actually makes much difference wrt happiness and satisfaction, thankfully.

[Evan Owen @EvanOwen]

@avidseries I think she’s highly intelligent, though somewhat naive, and pursuing a very successful career at the expense of what’s left of the integrity of broadcast journalism (not much).

[Taelin @VictorTaelin]

this is my personal singularity moment

this post may sound like a paid ad. I only wish. I'm concerned, more so than happy. the world is changing, and, among the scenarios where AI goes terribly wrong, inequality is the most realistic, yet, the one Anthropic seems to be the least concerned about. I'm glad OpenAI is taking the opposite stance: *personal AGI for everyone*. I think this is a commendable position in the times we live. but who am I in the queue of the bread?

anyway, Fable is here, so I'll just report my first-hour experience

first of all, all my pet prompts are solved.
→ λ-calculus puzzles
→ bug questions
→ one-shot apps
all are trivial to it.

I don't have anything harder other than my
ongoing work

so, in the last several days, I've been toying with HVM5, a new interaction net evaluator with a faster loop.

after writing the first version, I left 32 GPT-5 agents working for ~20 hours each. this resulted in up to 2x speedups, but the file size increased by 2-fold and quality decreased significantly.

I then simplified the whole thing into an even simpler core, and left Opus 4.8 and GPT 5.5 optimizing it for 8 hours. Opus got a legit 6% - 34% speedup in most benches. GPT got better results, but, sadly, an unusable file.

I then asked Fable to optimize it.

2 hours later, it landed a 1770% speedup in one case, 100%+ in other 4, and 22% in average. yes, in 2 hours it outperformed me, opus 4.8 and a swarm of gpt 5.5 agents, by one order of magnitude.

that could not possibly be legit. "it must be hardcoding the benchmarks" (GPT trauma). so I read its explanation and what it did was, indeed, the most high impact optimization one could try first. seems like HVM5 was wasting a lot of time garbage-collecting unused branches of pattern-match nodes. I had optimized that for static mats, but not for dynamic mats. skill issue. Fable figured how to do it for these, resulting in a massive speedup in some benches

but wait, is that *correct*? I'm not sure yet, it is credible, but this is the kind of thing that is very easy to get wrong on interaction nets. the problem is, when I was ready to start auditing Fable's solution so I could tell whether it was buggy or legit, it interrupted me to tell me it had found a massive bug on the code *I* had written.

... wait, what?

so... for garbage collection purposes, I stored a bit on lambda term pointers that meant "the variable bound by this lambda has been freed, so, its lambda must free whatever argument it is applied to". that's fine. yet, on duplicator nodes, I also used the same bit to mean "one of the duplicated variables was freed, so, treat this dup as a passthrough no-op". so, if a lambda entered a duplicator, it would mistake the lambda's collection bit for its own, resulting in corrupted interaction!

that's a mouthful, why I'm writing this?

just so you can appreciate the sheer absurdity of what just happened. I didn't ask it to find bugs. I asked it for an optimization. and even if I did ask it to find bugs, this bug is so astonishingly subtle and specific, identifying it takes mastering the domain to an extent that it beyond even me. I'd easily need hours or days to fix it, *if* I ever came across it. chances are it would just go unnoticed. and Fable found it and fixed it like it was nothing, while it was busy adding a 17x speedup to a file that neither I, nor Opus 4.8, nor a fleet of GPT 5.5 managed to barely make 2x faster.

oh and there is also another tab where it is also ripping through Bend's codebase and finishing everything I had to do

I don't know what to say anymore

this isn't about Anthropic or OpenAI, this is about our collective future as a species. the world is changing, and we need to be aware of it, and discuss how to handle this change.

receipt below . . .

[Evan Owen @EvanOwen]

@ZackKorman How are you feeling now that it’s out?

[Evan Owen @EvanOwen]

@robertcourson correct, they need a bigger base model:

[Evan Owen @EvanOwen]

@Miles_Brundage @AndrewCurran_ Generally the thing OAI needs to do is speed up successful pretrains. The RLHF team had to squeeze every bit of juice out of Orion through to 5.4 because Spud took so long to finish.

If Mythos is as far ahead of Spud as it appears, they need another new base model soon.

[Evan Owen @EvanOwen]

@xikhar @haider1 Where’s the evidence Fable is distilled? Benchmarks look stronger than the Mythos preview release and they explicitly say this is just Mythos with safeguards in place.

[Evan Owen @EvanOwen]

@adxtyahq right? can’t believe such a large model is actually cheaper than opus 4 was at launch

[Evan Owen @EvanOwen]

@Miles_Brundage @AndrewCurran_ Generally the thing OAI needs to do is speed up successful pretrains. The RLHF team had to squeeze every bit of juice out of Orion through to 5.4 because Spud took so long to finish.

If Mythos is as far ahead of Spud as it appears, they need another new base model soon.

[Evan Owen @EvanOwen]

@Miles_Brundage @AndrewCurran_ The problem is that 5.5 was the first Spud release, and it falls flat (even 5.5 Pro) compared to Fable 5.

My guess is that Mythos is a much bigger base model than Spud, so OAI have a lot of catch-up to do with 5.6.

[Evan Owen @EvanOwen]

@nickmacco Envious!! Looking like a fun week over there.

[Evan Owen @EvanOwen]

@jlongster yes. even 1px shifts can add an unconscious sense of jank for the least discerning users

another good rule of thumb is to only let layout change either (a) on a user interaction or (b) after loading state, and minimize the changing surface area to create a sense of stability