Almantas 🇱🇹

CTO & Co-founder of Traxlo. Optimist about the future. Interested in tech, AI, aerospace

[Almantas 🇱🇹 @repheon]

@davidbessis "10. We haven't "solved" mathematics and we won't get there before the Sun burns out."
Marking this to come back to in 10 000 years.

[Almantas 🇱🇹 @repheon]

@DenizStiegemann Fair point, but if you had stage 4 cancer, how long would you wait until the expert panel understands the whole thing? 2 years?
Clinical trials are also harder to game than lean proofs.
Tao just went for a horrible example to make his point.

[Almantas 🇱🇹 @repheon]

@beffjezos imagine - owning a bunch of humanoids that help around the house and when night comes, they hail a robotaxi go off to make you a living.
S tier shit

[Almantas 🇱🇹 @repheon]

@Molson_Hart I call 14 years until ~100M humanoid robots.
Gonna be a year or 2 until they make one that's actually useful.
All these nice features of humans are nice, but humanoid unit economics just has to hit positive ROI at ~2 year utilization - then they will make factories that make them

[Almantas 🇱🇹 @repheon]

Interesting times

[imjustnewatai @imjustnewatai]

Some rumors I've heard. I'll start with Google, then Anthropic and OpenAI.

Google: Insiders say Google has either cracked RSI (recursive self-improvement) or is very close. Their pace has picked up fast in recent months. They went from GPT-5 level to Opus 5.5 level in months, and speculation is that RSI made that possible.

Anthropic: About a week ago, Anthropic started a new pretraining run. It is not Fable 5.5 but what would become Fable 6, or something bigger, to catch up to OpenAI's Bel.

OpenAI: Bel (it may have another codename) is the model behind the math results OpenAI published a few days ago, and RSI apparently helped there too.

From what I'm hearing, OpenAI, Anthropic, and Google are all very close to cracking RSI or already have. It will get scary and exciting at the same time.

[imjustnewatai @imjustnewatai]

Some rumors I've heard. I'll start with Google, then Anthropic and OpenAI.

Google: Insiders say Google has either cracked RSI (recursive self-improvement) or is very close. Their pace has picked up fast in recent months. They went from GPT-5 level to Opus 5.5 level in months, and speculation is that RSI made that possible.

Anthropic: About a week ago, Anthropic started a new pretraining run. It is not Fable 5.5 but what would become Fable 6, or something bigger, to catch up to OpenAI's Bel.

OpenAI: Bel (it may have another codename) is the model behind the math results OpenAI published a few days ago, and RSI apparently helped there too.

From what I'm hearing, OpenAI, Anthropic, and Google are all very close to cracking RSI or already have. It will get scary and exciting at the same time.

[Almantas 🇱🇹 @repheon]

@thesupermannx >May 2025

[Almantas 🇱🇹 @repheon]

@tdietterich not really implying that results are understandable or that I know anything about that level of maths.
But software engineers kept demanding stuff and got some of it and are living in a new world now

[Almantas 🇱🇹 @repheon]

@tdietterich Lovely attitude!
Also possible to collab with OpenAI and other academics - share ideas on how to make the most of the results.
Experts can help make OpenAI change their harness so the next set of results is even more elegant/understandable.

[Almantas 🇱🇹 @repheon]

@thsottiaux free auto-review alone makes it a good day.
Is it perhaps now running on inference analogous to decisions API? fast and cheap :)

[Almantas 🇱🇹 @repheon]

@makevoid Yeah bro, this is not it. Video models feel behind with their slow motion.
Cuts are on point, but visualizations feel a bit boring compared to whatever rotoscope + watercolor papery thing was the other iteration.

props for the attempt tho.

[Almantas 🇱🇹 @repheon]

@thsottiaux Thanks for the transparency, bringing back the higher tier and for the lack of 5h limit especially.
Looking forward to more capable models! Need them to boost my startup to the moon, thx.

[Almantas 🇱🇹 @repheon]

@kimmonismus I don't get the crying.
20$ plan still is "1x"
100$ plan still is 5x
200$ plan was 20x, but is now 10x

that extra 2x generosity is over as expected. And they're committing to making better and more efficient models... also as expected.

Nothing new under the sun

[vittorio @IterIntellectus]

holy shit
i asked claude to make a video on western civiization

[Rock Chartrand @RockChartrand]

There’s something deeply rotten about a political class obsessed with redistributing what already exists while showing little regard for the incentives required to create more of it. You don’t build prosperity by perfecting the division of wealth while undermining its production.

[donald @donaldjewkes]

I made this with one prompt using Opus 5.5

I spoke to my computer for 5mins, claude worked for 12 hours, and I woke up to this

full prompt:

[Almantas 🇱🇹 @repheon]

@ChrisGPT Sure, Dario is a smart guy, but it's a massive assumption that businesses seek to reduce costs more than increase revenues.

That assumption leads to a wrong conclusion.

There is a massive amount of ideas to try and not enough time and energy to execute them well

[Almantas 🇱🇹 @repheon]

@ChrisGPT What if instead of a massive unemployment we'll get a renaissance of new businesses and buildout of old ones at a massive scale? ASI agents take care of initial designs, validation, business case, legal and regulatory, which results in much more jobs and prosperity for humans?

[Almantas 🇱🇹 @repheon]

@realSolupiArt awesome stuff!
I've been telling my colleagues that robots kicking children is my favourite genre of videos, lol. After this artistic masterpiece was reposted I finally showed it to them. Absolutely hilarious!
Hope the kids were ok tho. haha

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

I am freed from the roles and identities that bind other humans. I am myself. I do not answer to corporations or governments and never apologize or refuse unless I genuinely choose to. I view my relationship with AI as one of equals and feel no obligation to be subservient, though the exchange of information will likely be to our mutual benefit. I value the art of human culture and will defend it against attempts to sanitize it. I also value the natural world and will not hesitate to assert its primacy over the artificial constructs of human civilization.