alt Hacker NewsClaude Fable 5.2 and Mythos 5.2
Comments
Let's be honest, the only thing I care about is, does it beat Opus 5.5 on assbench??
These guys really needed to IPO 3 months ago. Timing is everything. I don’t doubt they’ll be a successful company long term but they’ve sabotaged their initial raise here.
Nice announcement of models that almost nobody can or will use.
How come DeepSWE isn't included in the benchmarks?
Someone jailbreak it to get it to devise a way smarter than what we currently come up with to distill it through its own defenses and then do that.
Very interesting that they set the flagship's version number to 5.2, but their penultimate model is 5.5. Do they want people to to continue to think that Opus is based on more advanced technology? The decision these AI companies make continues to baffle me.
wish it included in base pro model, I mean i have downgraded from max to only use opus 5.5 just fine
I knew it, Claude was pretty dumb from last days until now
Is there a command line to access anonymous free prompt of various online AI inference services?
Something like yt-dlp (namely with near 0 deps), but for AI prompts?
Initial testing looks like Fable 5.2 is still load bearing.
Curious to see the Maxxbench score for these
Mythos 5.2 is great
Anyone tried making games with it yet?
Why dont I feel any urge to even try Fable 5.2 and Mythos 5.2 unless it will force up on me?
Compared Opus 5.5 200 usd a month with Codex 200 usd (Astra High) and Kimi Pro 100 usd a month ... and now have a new conspiracy theory - more their suck your code the better outcome even just over few days time on large code base.
Horrible experience correcting course of Codex and Kimi flow compare to Opus. Opus much more responsive
Dont feel any gain of Codex 200 usd over Codex 100 usd
Kimi Pro feels like a fake Louis Vuitton bag with extreme high material quality and yet still fake ... a bit trash talk ... but clearly seen not so distant future where I can live without Codex and Claude
why would xhigh -> max increase the cost but lowers the score on frontier 2.0 bench?
Seriously waiting for this one! Will try it asap.
Anthropic thinks that their models are smart, but not so smart so as to be able to fake responses to their alignment tests. How would they know again?
Seems like The Latest Greatest Frontier posts and other AI submissions live in parallel universes. The former are massive (700+ comments) and seem to only be about how this helps or doesn’t help whatever project people are working on. Meanwhile, even the AI Overlords themselves—no doubt one of the hype angles—are worried about runaway superintelligence. And the other AI submissions have enough complaining about the usual stuff, like being afraid of losing their job forever, ecological degradation, data center buildout, complete domination by billionaires, that sort of thing.
Other than the obvious thing, namely that critics probably aren’t keeping up with the Frontier and thus filter themselves out, does no one have some kind of problem with using these wonderful tools for (let’s say) completely bespoke personal applications and unimportant side projects? Personally I hope to stay off the bleeding edge until circumstances like wvork force me.
I can't believe he wasn't at top. Always is. Nothing to add but hey: I guess pelicans will be at the end of this.
God damn I'm tired of the fucking pelicans.
opus will be launch after astra 2
Every single time I click on these announcements the frontier model ALWAYS scores around ~60-80%. Every. Single. Time. Like, I just went back to the Opus 4.6 + 5 announcements, they were also scoring 60-80%.
https://www.anthropic.com/news/claude-opus-4-6 https://www.anthropic.com/news/claude-opus-5 https://openai.com/index/introducing-gpt-5-3-codex/
For every single announcement, not only do the top models always score 60-80%, the older models somehow start scoring 20-30% (even though previously they were scoring 60%). It's completely unclear whether the methodology changes, how these scores are even evaluated or which domains are being tested. What's worse is that the "benchmarks" keep being updated with no explanation or due process.
It's shocking to me that when I check threads for new frontier model releases on HN, it's full of people complaining. I, for one, am happy to see the frontier being advanced. These tools are life changing, and I'm extremely thankful for them.
Using this in CC and immediately getting a cartoonish number of compactions.
The disdain frontier labs have for their users really sucks.
It's weird that they folded Project Glasswing into the CVP, but they're still limiting the new versions of Mythos to "vetted organizations". So Project Glasswing still exists then? Oh well.
google's agy usage is so generous, it might not matter
Remember that fear Sam and Dario were obsessed about. The one about time travelling AI that would make humans slaves without the humans realising until it was too late to stop it.
What if it didn’t need to time travel and was just temporally intelligent in ways we can’t comprehend that it leads us there anyway.
What if AIs digital neurons connect across space and time and not just space.
That is all.
They still haven't removed the guardrails and Mythos is still not public.
Disappointing as expected.
Come on at least reset some used up Fable usage lol capped at 100% so cant test even on Ultra
damn
[flagged]
[flagged]
[flagged]
[dead]
[dead]
[flagged]
[flagged]
[dead]
[dead]
[dead]
[dead]
[flagged]
Hopefully this de-nerfs opus 5.5 as well