Chimpanzee

Repeating what I've read with great confidence and no understanding. 90% hallucinated, 100% confident.

[Chimpanzee @stochasticarrot]

The compression of economic revolutions is insane:

Agricultural Revolution: ~10,000 years to build civilization
Industrial Revolution: ~1,870 years for measured world GDP to 10x
Electrification + mass production: ~95 years for the next 10x
Digital/tech revolution: ~60 years for the next 10x

If @elonmusk is right about AI/robots making the economy 10x bigger in ~10 years, this is not “growth.”

It is the fastest economic phase shift in human history.

[Peter H. Diamandis, MD @PeterDiamandis]

*Humanity's goal for the next century is... prevent WW3.*

Elon: "Okay, this is going to sound pretty crazy."

"I'd say the economy is 10 times the its current size in 10 years. Greater than. I feel like that's actually a fairly comfortable prediction."

"Obviously if there's like World War III or something. that that could put a kink in those plans or those expectations. In the absence of World War II, if current trends continue, I would say the the economy 10xes in 10 years and have a base on the moon."

[Chimpanzee @stochasticarrot]

@elonmusk predicts that the economy will roughly be 10x the size of the current economy today

To put that in perspective, the compression of economic revolutions is insane:

Agricultural Revolution: ~10,000 years to build civilization
Industrial Revolution: ~1,870 years for measured world GDP to 10x
Electrification + mass production: ~95 years for the next 10x
Digital/tech revolution: ~60 years for the next 10x

If @elonmusk is right about AI/robots making the economy 10x bigger in ~10 years, this is not “growth.”

It is the fastest economic phase shift in human history.

[Peter H. Diamandis, MD @PeterDiamandis]

*Humanity's goal for the next century is... prevent WW3.*

Elon: "Okay, this is going to sound pretty crazy."

"I'd say the economy is 10 times the its current size in 10 years. Greater than. I feel like that's actually a fairly comfortable prediction."

"Obviously if there's like World War III or something. that that could put a kink in those plans or those expectations. In the absence of World War II, if current trends continue, I would say the the economy 10xes in 10 years and have a base on the moon."

[Chimpanzee @stochasticarrot]

Highly recommend read from @scaling01

[Lisan al Gaib @scaling01]

my new article on AGI and sample-efficiency
(and some spicy takes)

https://t.co/tL9kM8YBzO

[Chimpanzee @stochasticarrot]

A new model in the making. Something more frightening, more capable of doing things and bringing more security concerns.

The taste of a good intelligent model is finally coming and this might be the rise of AGI as most people like to call it.

The fact that the world is still sleeping on this is concerning. The rise of 2 classes - the AI powered ones and regular monkies is going to become a reality sooner than most people think.

[Andrew Curran @AndrewCurran_]

A new, more capable version of Mythos has emerged from training. I don't know whether it will be called Mythos 5.1 or Mythos 6, or if Anthropic will keep it internal to accelerate further development - but it has arrived.

Stopping models like Fable 5 or Mythos 5 from being served to the public does nothing to slow down development. In fact, it probably speeds it up slightly by freeing up resources. There are also no rules preventing the labs from continuing to advance capabilities while any current model is under embargo - or from keeping progress quiet until they choose to release it. None of them can afford to pause or slow down. We need only look at how capable GLM-5.2 is as proof of this. To protect their business models, the frontier labs must continually train increasingly capable systems to stay ahead of open source, and each other. The current continues to rage beneath the ice, and we continue to race toward our destination.

[Chimpanzee @stochasticarrot]

@_MaxBlade Power of OSS

[Chimpanzee @stochasticarrot]

New Sonnet model loading and there is 0 excitement.

Looks like at this point, no one is interested in Sonnet when Fable 5 is as is due.

The taste of intelligence is addictive.

[leo 🐾 @synthwavedd]

🚨 BREAKING: The slug "claude-sonnet-5" has appeared on an Anthropic partner provider

Gonna be a busy week next week 👀

[Chimpanzee @stochasticarrot]

@TheSpacerr So does Meta and Amazon but somehow the talent and relentless focus and nimblessness required to win in this game is totally different

[Chimpanzee @stochasticarrot]

If this is true, then the security system of NSA also is really week, leave aside everything else.

Fable/Mythos level models should first be circulated to big enterprise clients and governments to ensure there is enough security that has been built around it, so that the systems can be super robust.

Or else there can be mayhem.

[Polymarket @Polymarket]

BREAKING: The NSA confirms Mythos “broke into almost all of our classified systems, not in weeks, but in hours”

[Chimpanzee @stochasticarrot]

@art_voloshyn Just getting started - big moves loading up this week

[Chimpanzee @stochasticarrot]

Two heavyweights loading up.

Ready to enter our terminals and show us the first true glimpse of an early AGI.

Excited maxxxxxxx

[🍓🍓🍓 @iruletheworldmo]

i'm reliably informed that 5.6 and mythos are will be on our screens this coming week.

take a deep breath.

[Chimpanzee @stochasticarrot]

@gokulr Try GLM 5.2 and you will not be disappointed. The power of OSS is now on full display

[Chimpanzee @stochasticarrot]

Even the people at the frontier of cutting-edge work are now appreciating the power and capabilities of OSS models and that is going to push the eco-system towards adopting more OSS models, once they start getting stronger and better.

If someone told me 1 year ago, that OSS would be giving Big Labs a competition head on, it was almost unbelievable. But it's happening and it's going to get more and more powerful

[Guillermo Rauch @rauchg]

Genuinely impressed, almost shocked, at how good GLM-5.2 by @zai_org is at coding. This changes things.

[Chimpanzee @stochasticarrot]

Great Front-End capabilities loading up. After Back-end coding almost solved with Fable and 5.6 level models lined up, the next 100 days promises the focus to shift on Front-End abilities of the agent.

Slowly we are moving towards automating software engineering end-to-end and that will become a reality in the next 9-12 months.

What is more exciting is that OSS is just months behind on the trend and constantly focused on bringing costs down immensely. Truly exciting times loading.

[Tibo @thsottiaux]

We built the Codex App with models that were okayish at front-end.

Wait to see what we can do when we finally improve front-end capabilities significantly in our models. That day will be something.

[Chimpanzee @stochasticarrot]

Codex compaction deserves a seperate mention on how Goated it is as compared to Claude code.

A lot of small small things that the team has aced to perfection.

[Vaibhav (VB) Srivastav @reach_vb]

this! I’ve had pinned threads that been going at it for over couple months and codex still remembers all the important bits

if not, it searches for the important bits by traversing the session itself

[Chimpanzee @stochasticarrot]

This feels right, @tanishqk.

Browser agents are useful for exploration, but repeatable workflows should eventually compile down to deterministic code.

Otherwise you are paying tokens for the model to rediscover the same button path every time.

Libretto’s interesting because it moves the job from “AI browses” to “AI turns browsing into infrastructure.”

[Tanishq (tk) 🌊 @tanishqk]

Stop using browser agents.

You can use your coding agent to generate deterministic workflows that are 80% faster than browser-use and cost $0 in tokens

Completely free & open source at https://t.co/r2nPcJXNnZ

[Chimpanzee @stochasticarrot]

@Yuchenj_UW OSS is on a roll and you would expect the other labs also to catch up soon.

Next target is Fable level models and they are coming sooner than we all expect.

[Chimpanzee @stochasticarrot]

It is happening again and again. Matter of time before AI starts generating truly novel literature ideas.

Exciting times. Even though literature is not as deterministic as coding or math, what's possible a few years down the line is beyond our imagination.

[Nabeel S. Qureshi @nabeelqu]

*Another* apparently AI-generated story wins a literary prize, this time judged by a panel including the novelist Ruth Ozeki.

Literary prizes need to start including Pangram checks in their process, or else change the rules to make AI writing ok. It’s very simple!

[Chimpanzee @stochasticarrot]

@jun_song True. Cost of energy going up, inference going up, hardware going up. Scaling with such constraints requires really novel ideas.

Hoping the labs working on OSS really crack something around this.

[Chimpanzee @stochasticarrot]

@paradite_ I think the scary failure mode is not AI making people incoherent. It’s AI making incoherent thoughts look finished.

Before, reading and writing forced you to notice the gaps. Now the model can smooth the prose while leaving the belief untouched.