knivesysl

the talisman of deep learning.

[knivesysl @knowrohit07]

mother of all demos. pt 2.

bodega OS is beautiful, unlimited and all ai models running locally on your hardware.

from porsche, westworld, to learning different planes of power, the giests and vishnu. playboi carti & personal notes at the end too.

it's crippling to use any other ai now. everything else feels bleak and lifeless.

[knivesysl @knowrohit07]

openai and anthropic will no longer exist by 2027.

recession and civil war is intertwined by 2027. so expect a big boil.

there won’t be highest tier intelligence, llms are just function aggregators,you can aggregate past a threshold. ( which we have reached ). it will be commoditised like a tts.

space x s1 prospectus if you read it then you wouldn’t give out brain dead opinions on it.

nvidia will survive. wont reach 10T

[Lisan al Gaib @scaling01]

Cold take on what comes next:
- OpenAI will flourish
- Anthropic will continue to be profitable
- Google will not catch up to Anthropic or OpenAI
- no chinese company will catch up to Anthropic or OpenAI
- the highest tier of intelligence will become a luxury product that only companies and multi-millionaires/billionaires can afford
- most of the companies that invested massively in them will have massive returns
- SpaceX’s AI will be fine and on par with Google by end of year
- Nvidia will become the first 10T company

[knivesysl @knowrohit07]

openai and anthropic will no longer exist by 2027.

recession and civil war is intertwined by 2027. so expect a big boil.

there won’t be highest tier intelligence, llms are just function aggregators,you can aggregate past a threshold. ( which we have reached ). it will be commoditised like a tts.

space x s1 prospectus if you read it then you wouldn’t give out brain dead opinions on it.

nvidia will survive. wont reach 10T

[knivesysl @knowrohit07]

@jino_rohit vllm APC is fine enough.

[knivesysl @knowrohit07]

@willdepue hunt for food.

[knivesysl @knowrohit07]

@SIGKITTEN @mweinbach lil bro. you have a p****y pfp😭

[knivesysl @knowrohit07]

@mweinbach @SIGKITTEN word! do you use your studios as well, running llms locally?

[knivesysl @knowrohit07]

@mweinbach @SIGKITTEN i have seen you multiple times providing support for codex stuff “only”

until and unless you really fw w codex a lot lmao.

[knivesysl @knowrohit07]

@mweinbach @SIGKITTEN are you paid to be their community support.

[knivesysl @knowrohit07]

@francoisfleuret @julien_c @_akhaliq @huggingface zoom. the mac top.

[knivesysl @knowrohit07]

@benln all deemed to be failed. if this is the signal they show, then they are fucked.

[knivesysl @knowrohit07]

i’m gonna open source the training recipe for training a drafter model. and for infernece workloads, here are the benches and all the inference techniques which cloud llms have but now on apple silicon. Bodega inference engine is the fastest runtime out here.

https://t.co/S5uHrV0Wie

[knivesysl @knowrohit07]

@PalantirTech slop

[knivesysl @knowrohit07]

@anemll is it a joke or you still believe speeds has anything to do with ssd and ram.

[knivesysl @knowrohit07]

@ivanfioravanti use one. you are welcome ivan.

https://t.co/k4I6YXry9a

[knivesysl @knowrohit07]

@ivanfioravanti like?

[knivesysl @knowrohit07]

@blaire_pang hmmm very off comparison.

[knivesysl @knowrohit07]

@blaire_pang it will make your shipping of any service or product more reliable than the ones others have mentioned due to unreliable uptime’s and unknown surcharges.

if you wanna know more hmu.

[knivesysl @knowrohit07]

@dsiroker perfect. you are welcome to our subreddit for Mac Studio.

https://t.co/HJBcuQvUCm

[knivesysl @knowrohit07]

@ryanvogel your welcome

here you go:

https://t.co/bgwdTkxO1c