MaxLongTop (@bebankless) header
MaxLongTop (@bebankless) profile picture

exploiting inefficiencies in global markets building VANGUARD - AI trading system @DigitalAssets

In the arena
Joined April 2013
Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)Photo by MaxLongTop (@bebankless)
dario says anthropic has to buy compute 1-2 years before it knows if revenue covers it. ipo's days away
1
42
if MMs are following the script laid out since last month BTC should go to 67-68 next
39
saylor bought 485 btc and $228m of his own preferred this week. the coins are the marketing budget
Strategy has acquired 485 $BTC and repurchased $228M of $STRC. As of 11/15/26, we hold 850,849 BTC and $4.9B of USD Assets. $MSTR strategy.com/press/strategy-…
28
Lo del Board de Nvidia es una autentica verguenza. Por ejemplo, que pinta Melissa que viene de toda la vida en Taco Bell!!? y ademas tiene la tia 1799 acciones!!! Cero skin in the game…votando para echar de Nvidia al fundador. Estamos locos… Asi con todo
30
79
6
661
103,519
jensen huang is discussing a comeback with allies, per ft. nvidia was a one man show and the board just found out
All @JensenHuang has to do is tweet that he wants to be reinstated as CEO and gears will start moving. If Jensen says he wants back in, all of Nvidia and indeed the global AI ecosystem is still wired around him. The coup has no legitimacy. The anti-Jensen board members will immediately step down or likely be legally forced out. Under Nvidia’s bylaws, shareholders with at least 15% of votes for one year can demand a special meeting to remove sitting directors (requiring a simple majority with cause, or a 2/3 supermajority without) and elect a new board. And anti-Jensen board members know they’ll be held responsible for $1T in market cap erased if they hold out against the Jensen restoration. But these are NPCs, so they’ll buckle quickly. Trump also realizes there’s no way that removing his top AI CEO, namely Jensen Huang, will somehow keep him on top in AI (sorry, SI). He can overrule any lieutenant who pushed too far in their attempt to do to Jensen Huang what Xi did to Jack Ma. This “spy” nonsense is just like Russiagate, the exact kind of fake story that Trump himself was subject to, so he gets it. After all, there’s really only a tiny minority of evil conspirators in media and politics who actually wanted Jensen Huang out of the company he built from scratch. Almost everyone else in tech, both inside and outside Nvidia, from billion dollar funds to unicorn founders, admires and respects one of the greatest CEOs of all time. Jensen just has to ask for his company back. And an army will rally behind him.
62
congress is writing itself a veto over chip deals trump already approved. bipartisanship, finally
31
173b wiped in 24h with 1.8b in liquidations. the rest just left
22
Everything else is trivial, this is the real reason they took him out
For my first post, I’m sharing a letter @NVIDIA signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. The world needs both frontier closed models and frontier open models. images.nvidia.com/pdf/Open-W…
1
35
My BTC PA prediction indicator doing serious work
3
41
nvidia used its own money to fund the startups buying its chips. the openai commitment got cut 250b to 125b
NVIDIA reportedly helped AI startups buy its own GPUs instead of lowering prices and selling its excess stock cheaper. According to a new report, former CEO Jensen Huang used NVIDIA’s money to fund AI startups and data centers so they could keep buying its chips. The problem is that NVIDIA had ordered far more chips and memory from TSMC than the market could actually use. CFO Colette Kress pushed back against these deals, even cutting a planned $250 billion commitment to OpenAI down to $125 billion. If true, NVIDIA may have been helping keep GPU prices high instead of letting excess supply make them cheaper.
60
chatgpt and gemini both shipped a wallet today. every cycle ends with a wallet
40
the eu has 27 members so tusk says it can be 27 times more productive. the arithmetic is the policy
Our next race is the AI race. By joining forces, Europe gains much more scale, power and influence. By coordinating our efforts, we can all be 27 times more productive, access 27 times more services and speed up our growth. This unified scale is exactly where the US and China are winning over us. First, let’s set up leading research institutes for AI. Second, let’s translate research into industrial capacity. Third, let’s make the 28 regime work. Let’s stop funding our ambitions separately, work together and secure Europe’s place in the AI revolution.
24
half a billion in banned blackwell and four chinese labs trained on it. the export ban is a suggestion
I have a major result to share. I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China. This sounds insane, but I discovered that almost every AI model carries a hidden signature of the GPU that trained it. When a model is first set up, the starting values of its weights are laid out in a pattern that depends on the exact GPU. Each combination of GPU and seed number produces a unique pattern—a sort of fingerprint of the hardware used to create the model. Traces of that pattern survive training into the published model. Again, this evidence isn’t based on any insider information or leaking. It’s based on a forensic analysis of the model weights released by the companies themselves. I searched more than six trillion combinations of chip layout and seed number and matched them to open-weights models. In doing so, I found conclusive evidence that six Chinese AI models—including DeepSeek-V4-Pro, Kimi K3, and MiniMax-M3—were initialized on NVIDIA Blackwell systems. Four of the models match Grace Blackwell, the GB200 and GB300 parts in NVIDIA’s flagship liquid-cooled rack-scale systems. I then verified these results on real hardware. Running unmodified PyTorch with the seeds I’d recovered, a Blackwell B300 reproduced DeepSeek-V4-Flash’s untouched starting weights bit for bit, and a rented GB200 reproduced Kimi K3’s to a correlation of 0.999995. A Hopper H100 reproduced none of DeepSeek-V4-Flash’s, and changing the seed by one broke the match. (DeepSeek’s seed is 953049583 by the way.) The method also was verified to match 24 known open-weights models to the hardware that developers said they used. It is exceedingly unlikely that these results are a coincidence. Across six trillion combinations, the best accidental match should reach only about 8 sigma (i.e., eight standard deviations above what random chance would produce). The real matches range from 49 to 1,385 sigma, and even the weakest single test hits 20. These were data-center-scale training runs. With some analysis of release schedules and model details, I estimate that the largest, Kimi K3 and DeepSeek-V4-Pro, used 2,000 to 3,000 Blackwell GPUs each. Clusters of this size cost on the order of $150 to $200 million each to buy, and across these labs alone would add up to more than half a billion dollars of export-controlled GPUs, not counting experimental compute or otherwise. It’s quite likely that we see multi-billion dollar expenditure on export-controlled GPUs. I am highly confident in this result, and it has been independently verified by senior researchers at OpenAI and Anthropic. You can read the full report at unemploymentlabs.com/checkpo…. There is much more I’ve identified about these models through other forms of checkpoint forensics like this, and will have more to share soon.
31
altman told 50 samsung presidents ai demand is double counted. only two labs drive the growth and he runs one
1
2
57
united will ban ai agents from booking its flights. the same booking typed by hand at 2am is still legal
36
$10b on one model family and anthropic told the sell side the market is winner take all. at that bill, it has to be
Interesting remarks Anthropic made in a meeting with the sell side: - Anthropic has the stability and the better talent that open source lacks. - The AI market is winner take all. As long as we execute properly, we will take all of it. - The contract with SpaceX will not be extended beyond the middle of next year. - Developing Opus 5.5 cost more than $10 billion.
Readers added context they thought people might want to know
The remarks are unverified second-hand notes from a Barclays analyst call on Anthropic meetings, not official statements. Poster corrected $10B cost to full Claude 5.5 family; filings show SpaceX deals extend to 2029 but terminable after early 2027 on 90 days notice. x.com/jukan05/status… fourweekmba.com/ai-anthropics-… sec.gov/Archives/edgar…
40
the center for ai safety warns ai is closing in on all remote work. claude finishes 41.3% of real freelance gigs, up from 2.5% a year ago
39
microsoft urged trump to scrap the planned ban on chinese transceivers in ai data centers. the buildout comes first
26
anthropic plans to ipo next week, wsj says. the leak is the roadshow
1
63
trump's best ai regulation so far is telling a ceo to say goodbye to his wife and children. the guy totally changed everything
39