Joshua Dyess

6.2K posts

Joshua Dyess banner
Joshua Dyess

Joshua Dyess

@JGCole234

Husband, Father, Catholic.

Brandon, MS Katılım Kasım 2009
855 Takip Edilen119 Takipçiler
Patrick McGee
Patrick McGee@wpmcgee·
Anyone that stood up and cheered for the ULM starter after he went 8.1 innings should be permanently banned from PTP.
English
30
13
128
52.2K
Joshua Dyess retweetledi
Milk Road AI
Milk Road AI@MilkRoadAI·
Leopold Aschenbrenner, the 24 year old who wrote a 165 page AGI manifesto, got it right on the money, and turned it into a $5.5 billion hedge fund. And he's identifying the single most important milestone to watch for in all of AI. The question is can AI automate AI research itself? Here's why that question matters so much. Right now, a few thousand human researchers at the frontier labs are driving all the progress. They design experiments, write papers, propose architectural improvements, build the next generation of models and it's an incredibly small workforce doing incredibly high-leverage work. If an AI system can do that job even partially, the feedback loop changes completely. The AI makes algorithmic improvements, which produces more powerful AI, which makes better improvements, faster. You go from linear progress to compounding returns and a decade of research could compress into a year. Aschenbrenner says there's a "pretty reasonable chance" this happens within five years. He's not alone, Anthropic says they're on track to fully automate AI R&D as soon as early 2027. OpenAI has publicly targeted a fully autonomous AI researcher by March 2028 and Sam Altman has said a research intern level AI will exist before the end of this year. If he's right, the next few years won't look like the last few years but they'll look like nothing we've seen before. The future is bright!
Milk Road AI@MilkRoadAI

Leopold Aschenbrenner, former OpenAI safety researcher, fired at 22, now running one of the best performing hedge funds on the planet. He grew Situational Awareness LP from $225 million to $5.5 billion in roughly one year beating the S&P 500 by 47% in just his first six months. And the core bet wasn't Nvidia or OpenAI, it was electricity. Here's his actual mat and it's worth reading carefully (Save this). GPT-4 trained on roughly 10 megawatts of compute, a100-megawatt cluster followed in 2024. By 2026, the leading training cluster is approximately 1 gigawatt, the power of the Hoover Dam. By 2028, 10 gigawatts more power than most U.S. states. And by 2030, a single training run could hit 100 gigawatts and consume over 20% of all US electricity production. That's just the training cluster, Inference GPUs on top of that. US power production has barely grown in a decade. He turned that observation into his biggest position, Bloom Energy, a fuel cell company that generates power directly at data centers without touching the grid. He accumulated the $875 million position through 2025 while shares were in the mid teens and that position is now worth nearly more than $2.2 billion. The logic is that everyone is racing to build the most powerful models in history but those models require power that doesn't exist yet on the grid. The companies that control where the electricity comes from, that's where the leverage is. His fund now holds Bloom Energy, CoreWeave, Core Scientific, Lumentum, and a short on Infosys, the last one a bet that AI will gut traditional IT outsourcing. We at Milk Road have been saying it for months, the picks and shovels play in this AI cycle isn't the chip companies, it's the infrastructure keeping them running. Come join us so you can see our entire thesis. Link in bio!

English
30
116
1.5K
576.9K
Joshua Dyess retweetledi
Matthew Yglesias
Matthew Yglesias@mattyglesias·
A quirk of American life is that most of the people who sincerely enjoy urban living also have progressive ideological commitments that make them skeptical of pro-business politics, but those urban landscapes are the physical manifestation of large-scale capital investment.
English
11
80
870
43.8K
Joshua Dyess retweetledi
Paul Graham
Paul Graham@paulg·
This graph shows one of the many reasons the good old days were not so good. Even as late as 1870, women were losing 2 out of 5 of their children.
Paul Graham tweet media
English
62
86
908
81.5K
Joshua Dyess retweetledi
Paul Graham
Paul Graham@paulg·
@cb_doge I've already explained that this "quote" from me is bullshit. But the really striking thing here is how similar this tweet is to other ones appearing just as Musk v. Altman is happening. I smell a coordinated PR campaign, and I'm not going to be used as part of it.
English
44
43
1.2K
33.5K
Joshua Dyess retweetledi
Konstantine Buhler
Konstantine Buhler@Konstantine·
Sir @demishassabis has a mind for synthesis. His favorite book is about a grand theory of everything. His preferred philosophers are seen by some as opposites. His life's work ranges from board games to Nobel-winning science. We're grateful to have hosted Demis and his @GoogleDeepMind team at @sequoia AI Ascent last week for a fireside chat. He kindly gave us permission to share this, and you can watch the full video here: 00:00 Intro 00:38 The Common Thread 01:29 Games as AI Training 02:59 Startup Advice 1.0 04:39 Founding DeepMind 07:25 DeepMind and AGI 08:52 AI for Science 10:37 Biology Breakthroughs and Isomorphic 12:42 New Sciences 20:29 Philosophy
English
41
198
1.5K
400.4K
Joshua Dyess retweetledi
Robert Pondiscio
Robert Pondiscio@rpondiscio·
Yesterday I ordered an obscure book and a 20 pound bag of dog food on Amazon. Today it was delivered to my house in rural upstate New York. Remind me again why Jeff Bezos doesn’t deserve to be a billionaire?
English
265
361
12.9K
498.1K
Joshua Dyess
Joshua Dyess@JGCole234·
@rayfp I was there with my wife who knows very little about Bob. “Why is he dressed like a school shooter? Why doesn’t he talk to the crowd? Is he autistic? Why does he walk like ET?”
English
0
0
5
224
Ray Padgett
Ray Padgett@rayfp·
new hoodie-Bob applauding gif just dropped
GIF
English
7
44
441
14.7K
Joshua Dyess retweetledi
roon
roon@tszzl·
the concept of “robot taxes” is abhorrent even in principle. it’s not at all clear that owners of robots or ai services will be monopolistic or rather just abundant and then you are taxing a useful production commodity like a “steel tax” or “grain tax”. tax capital income instead
English
120
25
782
66K
Joshua Dyess retweetledi
Gavin
Gavin@PrimaryCinema·
The Man Who Wasn’t There, A Serious Man and Inside Llewyn Davis are 3 Coen Bros. movies that end with the introduction of an outside force so far beyond our comprehension, the struggles up to that point pale in comparison. Those are, an alien spacecraft, a tornado, and Bob Dylan.
English
25
271
4.3K
125.4K
Crepe Supreme
Crepe Supreme@crepesupreme·
@Ananyo Buried at the end: Tao saying the problem was 'maybe easier than expected... like there was some kind of mental block.' A 23-year-old without a Ph.D. did the trying-weird-things labor. ChatGPT Pro lowered the prior on which approach was worth checking.
English
1
3
119
21.4K
Ananyo Bhattacharya
Ananyo Bhattacharya@Ananyo·
23 years old with no advanced mathematics training solves Erdős problem with ChatGPT Pro. "What’s beginning to emerge is that the problem was maybe easier than expected, and it was like there was some kind of mental block.”-Terence Tao scientificamerican.com/article/amateu…
English
54
299
2.8K
1.2M