libpol @libpol_org

It'll be interesting to look at this tweet in 2028

Henrique Berthier @HBerthier

17 hours ago

@ChrissGPT Thats a good thing actually, we dont need fable class. We actually need is to have 5.5 class with few improvements and lowering required compute power. Pursuing endlessly the next frontier is doing more harm than good. We already have enough production-level models. What we need

0 0 0 410 0

@oldeborg @souljagoyteller I would bet anything that Wanye and Zero HP Lovecraft do, without a doubt

@tgstk714 @janecoaston Much like democracy is the worst system of government besides all of the others that have been tried, the only thing worse than universal suffrage is stratified suffrage.

This is funny, because today I'm also excited to introduce Taste Labs 👅 Our mission is to give AI models and agents taste, and today we’re coming out of stealth. AI has nailed objective domains and made it easy to generate anything. But it still feels off. Now, the challenge is taste. What's sweet, what's sour, what tastes like your mom's cooking, what’s DELICIOUS? This requires turning a fuzzy, subjective sense into something we can digitize and train on. There are two sides to cracking this, the embodiment layer and the model layer: - We’ve already been working with top tongue roboticists to design and build a cutting edge, robotic tongue, with over 20,000 artificial tastebuds—more than twice as many as human "supertasters." - We’ve also been working with model-layer companies to build out this new taste modality. - Finally, we're exploring new user-modeling opportunities and new modes of human-computer interaction opened up by robust AI taste technology. We want a future where AI doesn't just have taste—it can taste you. 👁️👅👁️ If you’re passionate about this mission, join us! (We're also looking for trademark infringement attorneys - DMs open.)

Thais Castello Branco @thaiscbranco_

19 hours ago

We’re excited to introduce Taste Labs. Our mission is to end AI slop. We’re building the data and infrastructure layer to give AI models and agents taste. And today we’re coming out of stealth, announcing our $18.5M seed funding, co-led by @CRV and @AmplifyPartners AI has

372 153 3K 1.0M 2K

@TexanTrojan94 @emmma_camp_ I'm not a libertarian and most seem pretty insane, but actual libertarian philosophy just genuinely isn't at all compatible with restricting voting based on sex/race/religion. It's those people who are faux-libertarians.

@xwanyex No, I think you're doing something much stupider, because I think you're actually being honest.

@aglassbrightly1 @janecoaston Why should we permit you to have the vote when you're sociopathic? Much to consider

ADHD “sufferers” be like yeah I finally started doing well at school after they gave me performance-enhancing drugs and doubled the amount of time I had for tests

@io_asc @zdrks @tenobrus to a degree, but overall it's palpably less grating. it's more like "oh there's that silly claudeism again", vs. with opus 4.8 where sometimes almost every sentence it writes in a turn is packed with claudeslop fluff

@rex_the_reptile @EricLevitz I remember him tweeting "liberals think they're your moral betters" in one of the classic angry follow-ups reacting to one of the dozen other times something like this happened. Correct, they do read your tweets and think that. You're quite sharp.

@internalazimuth @PositivistWitch I'm literally a normal dude from the year 2. It's you all who've gone off the rails.

@PositivistWitch The funniest and most pathetic part is I actually 100% believe he's genuinely unironically offended

@nosilverv seems like there is possibly some kind of mania-autism hybrid situation making manic autistic people particularly susceptible to a charismatic manic-autist psychosexual predator cult leader (that's my vague takeaway after spending like 10 seconds seeing a few tweets on it)

If one assumes the two orthodox factions are the Yudkowsky-likes (...one might even call him "ultra-orthodox", which I swear is not me trying to make a racist joke) and the Scott Alexander-likes, then former yes, latter no. Scott is deeply concerned but as far as I know does not have significant negative feelings towards them or their employees (if excluding him calling them all the antichrist: astralcodexten.com/p/my-antichris…) Most of the true doomer rationalists do think Anthropic (and OpenAI) are bad and in many cases completely evil, though, yes. Which I guess would be the rational view to hold if one has > 90% confidence a frontier AI lab will create a thing that kills all humans in our lifetime.

@abyssalblue_ @dumbfook myriad such circumstances

@OGTaylorB @tracewoodgrains @corsaren Liberals never realize when they're doing "the Omelas fallacy", refusing to even allow the counterfactual thought to enter their minds and reckon with the plausible upside and neutral accounting of opportunity cost

@tenobrus Good pushback — the solid reading: your instinct is *exactly* correct. The honest caveat is the one load-bearing flag from before; that's what's worth slowing down on. It's the genuine ambiguity we kept hitting, and I won't pretend it's resolved. It's *foundational*, not trivial.

I am pro-free enterprise and do not support state ownership of private corporations. That said, AI may, within the decade, become so crucial to not just national security but everything in general that all superpower nations' governments will have no choice but to have the government either partly own or at least have oversight power over the frontier labs. And anyway, the authoritarianism I mention from the Trump administration is like < 1% the AI stuff. It's everything, everywhere, all the time.

fable is really more like a manic pixie dream girl who swept into our lives promising to stir things up and teach us how to really live but then she got hit by a bus

Liminal Warmth ❤️‍🔥 @liminal_warmth

21 hours ago

The degree to which everyone is openly displaying frustrated anxiety at losing Fable after only four days of use is kind of interesting when you step back and think about it—junky vibes (me too)

9 1 175 17K 7

Gonna plant a flag that I believe AI will be the main political issue in the 2028 race