otecart

280 posts

otecart

otecart

@otecart

Software Developer. Love building stuff and tinker with AI (ML included)

Katılım Mayıs 2025
8 Takip Edilen1 Takipçiler
Elon Musk
Elon Musk@elonmusk·
@TrungTPhan The thinking tools of physics are incredibly powerful and can be applied to any field
English
265
262
2.9K
60.3K
Trung Phan
Trung Phan@TrungTPhan·
Elon crunching the cost to manufacture a rocket in a spreadsheet during a plane ride in 2002 and realizing SpaceX could do it at 1/50th the cost is the most productive use of a spreadsheet on a plane ride ever
Trung Phan tweet media
Trung Phan@TrungTPhan

SpaceX S-1 dropped and its centred on 3 business lines: ◻️ LAUNCH: 650 launches and 80% of global mass to orbit in 2025 ◻️ CONNECTIVITY: 9,600 StarLink satellites, 10m subs and 75% of maneuverable space satellites ◻️ AI: 1GW of compute, 550m active users Sales of $18.7B in 2025 (vs. $14B in 2024 and $10B in 2023). Operating loss of $2.6B in 2025. More forward looking, Anthropic is paying. xAI $1.25B a month to use Colossus for compute (!!!) through May 2029: "Pursuant to these agreements, the customer has agreed to pay us $1.25 billion per month through May 2029, with capacity ramping in May and June 2026 at a reduced fee. The agreements may be terminated by either party upon 90 days’ notice. The customer will retain ownership and intellectual property rights in its content, AI models, and related data. This structure allows us to monetize unused compute capacity in our infrastructure, while still permitting reallocation of the capacity for our own internal initiatives if needed in the future. We have sufficient capacity to provide compute for our own AI models, including support of our training and inference demands, and to satisfy the obligations under these agreements. We expect to enter into additional similar services contracts." Highlights its competitive advantage as a "repeatable, engineering framework": ◻️ Leverage our unparalleled launch capabilities to enable massive scale; ◻️ Identify and create new trillion-dollar market opportunities; ◻️ Design a solution with world-class engineering and first-principles thinking; ◻️ Apply “The Algorithm” (make less dumb, delete, optimize, accelerate, automate); ◻️ Vertically integrate all the way to the end customer; ◻️ Continuously drive cost down and throughput up; ◻️ Generate significant cash flow and reinvest in the future. The market opportunity page says TAM is $28.5 trillion: "We believe we have identified the largest actionable total addressable market (“TAM”) in human history. We estimate that our quantifiable TAM is $28.5 trillion, consisting of $370 billion in Space from space-enabled solutions; $1.6 trillion in Connectivity across $870 billion in Starlink Broadband and $740 billion in Starlink Mobile as well as additional opportunities in enterprise and government; $26.5 trillion in AI across $2.4 trillion in AI infrastructure, $760 billion in consumer subscriptions, $600 billion in digital advertising, and $22.7 trillion in enterprise applications."

English
26
85
1K
128.9K
otecart
otecart@otecart·
@fosaikt @TheLucasPg I wouldn't be so quick on turning Catholics down if I were you. Just remember how many Boeing whistleblowers died last years.
English
0
0
0
3
certamente não é o Lucas
Quando me perguntam pq a aviacao é segura eu falo que é porque nao tem jeitinho Ontem um Airbus A321 ficou impedido de voar e deixou seus 210 passageiros no chao, o motivo? Esse pequeno amassado de pouco menos de 20cm Em quase nenhuma outra área isso aconteceria
certamente não é o Lucas tweet media
Português
48
75
6.7K
481.3K
otecart
otecart@otecart·
@TheLucasPg @bcsanches Didn't notice you said "there is no room for cutting corners except when you have a tape"
English
1
0
0
1.3K
math
math@mouthbreath3r_·
@TheLaurenChen also meals in Asia are smaller and tend to have less meat and more filler carbs
English
2
1
16
362
Rational Aussie
Rational Aussie@rationalaussie·
@deredleritt3r This is literally what intelligence is - the ability to create accurate world models by testing different combinations of ideas and then verifying them against reality. The whole 'it was in the training data' argument never made any sense to me.
English
2
1
24
2K
prinz
prinz@deredleritt3r·
"The model didn’t 'invent' any 'new mathematics', [but] merely being able to know deeply all the results in a scientific field, and being able to use all known arguments expertly and with just the right choice of parameters, that alone can lead to a ton of breakthroughs, and this is not just limited to mathematics, this type of (extremely) solid expert execution is the bread and butter of many many scientific advances."
Sebastien Bubeck@SebastienBubeck

x.com/i/article/2057…

English
29
104
1.3K
136.2K
otecart
otecart@otecart·
@SebastienBubeck "Claim: AI can make scientific breakthroughs" Was it ever a question? Even a deterministic tool can do that, it depends on a task. The question was "Can AI create something new, something we never thought about?"
English
0
0
0
3
Ryan Els
Ryan Els@ryanels·
@viii_fn It used to be the default name 🤷🏻‍♂️
English
1
0
1
56
Ryan Els
Ryan Els@ryanels·
Do you still use master as your default branch? 🤔
Ryan Els tweet media
English
47
6
102
212.6K
Ryan Els
Ryan Els@ryanels·
@beausolai Yes. Besides who says they will ever even see the repo 🤷🏻‍♂️
English
1
0
0
386
Yuri Kushch
Yuri Kushch@YuriKushch·
@bevedoni @karpathy xAI gave lots of capacity to Anthropic. Also, I’m pretty sure that it’s cool to work on what you’re using on daily basis.
English
1
0
0
1.1K
Doniyor
Doniyor@bevedoni·
maybe dumb question, but why didn't @karpathy join xAI?
English
157
3
499
575.7K
otecart
otecart@otecart·
@thdxr I was thinking about using different agents (i.e. contexts) for writing tests and writing code, so that the agent who writes tests wouldn't have access to actual code, just functions' signatures, descriptions and behavior it needs to test. Haven't tried it yet though.
English
0
0
0
1
dax
dax@thdxr·
the latest flavor of agents being bad at writing tests is instead of testing a module, they do a dummy re-implementation of it and call that wonder what's making them particularly weak at this
English
28
0
167
16.5K
lossybrain
lossybrain@lossybrain·
@thdxr Are they writing bad tests before or after the subject under test.
English
1
0
0
418
otecart
otecart@otecart·
@NewAgain_11 @MattyWatty41 @22Hater_ People don't really lose their muscles much. Take a look at 40-ish yo men who were doing sport at school and then stopped. They are still bigger and stronger than average person even if they don't exercise for 10+ years.
English
0
0
0
29
NewAgain
NewAgain@NewAgain_11·
@MattyWatty41 @22Hater_ if that were true ppl wouldn’t lose muscle unless they were starving… stop training n eating high protein n see what happens
English
1
0
4
1.2K
UtdSetzzz
UtdSetzzz@UtdSetzz·
@ntajayb @WxnXBT @22Hater_ The guy i replied to was laughing at "body eats the muscle" Which it does My reply was in that context and I hadn't seen what he'd quoted
English
1
0
1
575
otecart
otecart@otecart·
@carllerche Go is ugly. Nobody likes ugly, even language models.
English
0
0
0
1
Carl Lerche
Carl Lerche@carllerche·
Which programming languages will thrive the most in the agent-era? Why will it not be Go? It is simple, consistent, predictable, and has excellent local reasoning.
English
47
2
61
15.6K
otecart
otecart@otecart·
@WakandaNonsens2 @brintypears I've seen some people reading really good books with the goal of reading these books for the sake of reading. They didn't get much from it. So I'd say it depends on how you read and what for, not just on what you read.
English
0
0
0
14
Rebecca
Rebecca@brintypears·
Not convinced reading is good for you based on people who read all the time
English
35
137
2.3K
207.3K
Kelvin Fichter
Kelvin Fichter@kelvinfichter·
Honestly the craziest thing about the GitHub hack is that someone at GitHub was still using VS Code
English
24
10
525
30.9K