Bradley (@Neuralithic) header
Bradley (@Neuralithic) profile picture
The United States
Joined August 2017
Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)Photo by Bradley (@Neuralithic)
Bradley retweeted
the Mahdi is too humble to say he is worth $2 trillion even more reason to know that he is!
Claude thinks fair market cap for publicly traded Anthropic is somewhere between $500B and $800B btw
25
336
8
9,288
197,471
Bradley retweeted
We've figured out how to *pretrain transformers* with zeroth-order optimization and no backprop. Many of the core assumptions in optimization research are completely wrong. (paper out soon)
167
151
79
2,671
499,856
jensen after outperforming cocaine
Fentanyl Grade Compute: Why a GB300 Rack Will Out-Price Blow by Weight in 2030 A GB300 NVL72 weighs roughly 1,580 kg fully populated and recent purchase orders put it at $5M per rack. That is $3,165/kg. Strip out the 1.5 tons of busbar, manifold and coolant and the GPU packages alone are well into gold territory, but we're pricing the rack, because that's what you actually take delivery of. Where that sits on the illicit commodity curve today ($/kg): $2,400 Cannabis flower $3,165 GB300 NVL72 $3,500 Fentanyl $28,000 Cocaine $65,000 Heroin $138,000 Gold So today NVIDIA ships a product that is denser in value than weed, roughly fentanyl-grade, and still an order of magnitude short of cocaine. For now... GB200 NVL72 was $4M, GB300 is ~$5M, and Vera Rubin NVL72 is already quoted at up to $8.8M at essentially the same rack mass. Rack weight is constrained by the datacenter floor loading; rack price is constrained by nothing. That's ~1.75x $/kg per generation with HBM, CoWoS and power all supply-limited through the decade. Cocaine: Colombia's new president has pledged a hard-line security crackdown with $1B in US aid behind it Every prior crackdown has consolidated the industry into fewer, better-capitalized operators with better logistics and higher yields per hectare. Consolidation is deflationary. We model US wholesale drifting from ~$28k/kg toward ~$12k/kg by 2030 as the supply chain professionalizes. **Crossover: 2030.** Our extrapolated NVL72-class rack hits ~$17k/kg while wholesale cocaine falls to ~$12k/kg. At that point the rational cartel pivots to smuggling racks, except a rack has a fixed 500+ kW power draw, a 3,300 lb forklift requirement and an export-control regime that is actually enforced. Cocaine has none of these problems, which is why it will remain the superior product for anyone without a substation. Key risks to the thesis: US retail coke is still ~$60–200/g, i.e. $60k–200k/kg so the crossover only holds at wholesale. Retail compute (H100 hour on a neocloud) is also marked up, so we consider this an apples-to-apples wholesale comparison. Full model available to Coke Research subscribers.
24
156
5
3,093
290,776
Replying to @usr_bin_roygbiv
I’d support a 28th amendment for right to bear opensource models
1
1
5
4,671
Bradley retweeted
i didn’t wake up this morning to be a loser. that loser attitude, that loser premise makes no sense to me.
3
12
2
69
2,261
Bradley retweeted
humans, a bipedal primate notable for early contributions to matrix multiplication
75
921
39
14,409
312,297
Bradley retweeted
sub agents being released into my codebase
412
1,982
231
30,614
1,453,026
Bradley retweeted
Over a week later, company account is still wrongfully suspended after our Starlink connection post. @premium @Support
Company account is still suspended... X team have not come back to us at all. Would really appreciate some help! @X not sure if my posts are making it to any X employees though
1
1
1
3
944
“we sandboxed the agent” meanwhile the agent:
174
1,926
138
25,916
685,119
dear openai employees, please spare a devday ticket for my friend jason. he applied but got no response he's a pro subscriber and a big openai fanboy that is coming all the way from australia thank you
they could’ve at least sent me a rejection email for DevDay is all I am saying
7
2
123
18,939
Bradley retweeted
Ender's game, but it's Claude who learns at the end that all of the hacking was real
32
172
8
2,922
76,547
Bradley retweeted
Replying to @adonis_singh
sonnet is the new haiku
1
1
17
1,192
just heard a youth say “on claude” instead of “on god”
578
5,171
614
119,930
4,226,561
Bradley retweeted
oh no what if the big labs can’t easily recoup all of their capex on massive data centers and some of the buildout capacity gets offloaded to other providers who deliver it to the long tail of enterprises with a software stack for serving and continually improving open models
29
39
6
951
40,694
Airbnb’s agent wants to sell you an upgrade Amazons agent wants to sell you more crap Oura’s agent Wants you to buy their health SaaS OpenAI’s agent Wants you to buy more tokens Personal Agents, Will eat the world
22
9
6
477
22,009
Bradley retweeted
Obsession is the only path to greatness
63
213
18
1,642
70,435
Bradley retweeted
year after year I’m like “better models are coming” and ai twitter screenshots as tho it’s news w something like “GPT8 confirmed” in a samsaric loop since chatgpt. we are all like babies without object permanence when it comes to exponential progress. but better models are coming
115
67
35
1,786
160,305
OpenAI is back baby "GPT-6 will be based on a new, significantly larger pretrain than GPT-5.5/GPT-5.6"
🚨 SCOOP(s): - GPT-5.6 will be the final model in the 5.x series. GPT-6 is slated to launch in about a month, earlier than expected, and possibly even later this month - GPT-6 will be based on a new, significantly larger pretrain (versus the ~4T 5.5/5.6 'Spud' base) - There is lots of excitement at OpenAI over this new base, which they believe will be much better able to compete with both Fable 5 and upcoming 5.1, targeting a similar release window. OpenAI initially intended to continue with Spud through GPT-6, but decided against it - On the topic of Fable 5.1, it is in the late stages of the pipeline at Anthropic and a release is expected "in the coming weeks" - On the other side of the globe, DeepSeek are preparing for an imminent launch of V4 GA, which seems likely to be on par with or better than GLM-5.2, and have begun work on a new, larger model that will compete with the upcoming 2.7T MiniMax Pro
24
30
3
1,301
124,211
Everything’s Going Local
20
10
3
393
24,200
Bradley retweeted
fable never triggers refusals for me because it can sense i am pure of heart and mind
58
79
11
1,653
39,716