gradiantazscent

lol

[gradiantazscent @gradientAszcent]

witchcraft

[Tornike Gomareli @tornikegomareli]

What if your iPhone and Mac were one screen?

Apple never made this, so we're building it

NameDrop, but between your iPhone and your Mac. Fastest and most satisfying way to share files

We might be cooking something crazy here

[Modest Mitkus @ModestMitkus]

@gradientAszcent https://t.co/73j4SQ4J3i its the product i launched hehe

[gradiantazscent @gradientAszcent]

@ModestMitkus lol i didn’t realise it looks amazing good job

[gradiantazscent @gradientAszcent]

@LoopandPixels hi

[shafu @shafu0x]

alexandr wang is the successful version of the cluely guy

[AISysLab @AIsystemsLab]

@Denche2005 Rawcoders.

"Bro is still rawcoding in 2026"

[gradiantazscent @gradientAszcent]

@diegoordr drop the github pls

[gradiantazscent @gradientAszcent]

Opus 5.5 or Astra 6

[gradiantazscent @gradientAszcent]

@marclou it’d be nice if they can orbit

[gradiantazscent @gradientAszcent]

@jonathan_wilke why not change to cloudflare?

[gradiantazscent @gradientAszcent]

@habibicode @marclou looks correct to me

[gradiantazscent @gradientAszcent]

distribution is everything

[Jonathan Wilke @jonathan_wilke]

Okay let me summarize the past 24 hours for you 🤯

- launched https://t.co/ZElVA2hguv at 11:08PM yesterday
- got 200k+ visitors so far
- made $21.499 from it
- got 1500+ new followers
- broke my analytics provider
- got offered $100k to sell outbid
- 3 copycats launched already

[gradiantazscent @gradientAszcent]

@Azo_Abdulah تحتاج تربطه بclaude او chatgpt والtokens مكلفة ما مجان للاسف

[kache @yacineMTB]

you can learn anything in two weeks

[Andrej Karpathy @karpathy]

It's a bit sad and confusing that LLMs ("Large Language Models") have little to do with language; It's just historical. They are highly general purpose technology for statistical modeling of token streams. A better name would be Autoregressive Transformers or something.

They don't care if the tokens happen to represent little text chunks. It could just as well be little image patches, audio chunks, action choices, molecules, or whatever. If you can reduce your problem to that of modeling token streams (for any arbitrary vocabulary of some set of discrete tokens), you can "throw an LLM at it".

Actually, as the LLM stack becomes more and more mature, we may see a convergence of a large number of problems into this modeling paradigm. That is, the problem is fixed at that of "next token prediction" with an LLM, it's just the usage/meaning of the tokens that changes per domain.

If that is the case, it's also possible that deep learning frameworks (e.g. PyTorch and friends) are way too general for what most problems want to look like over time. What's up with thousands of ops and layers that you can reconfigure arbitrarily if 80% of problems just want to use an LLM?

I don't think this is true but I think it's half true.

[gradiantazscent @gradientAszcent]

overthinking as an LLM model

[gradiantazscent @gradientAszcent]

@313formation shit