与「Random」相关的搜索结果

$BTC Finally reached $66,666. 👍 As promised, I want to change someone's holiday and send 0.15 $BTC(~$10,000) to one person by tomorrow. Just like, retweet and comment '$BTC'. Random winner in 24 hours. I will announce as a tweet at 12pm EST Sat

Giving away 1 $BTC to a random person! As promised, I’m sending 1 BTC (~$69,162) to one lucky follower. Like, retweet and comment ‘BTC’ to enter. * If you’re not following, you will be disqualified. Winner announced in 24h! Comments will be locked after 24 hours, the winner will be announced. Good luck everyone!

$BTC Finally reached $66,666. 👍 As promised, I want to change someone's holiday and send 0.15 $BTC(~$10,000) to one person by tomorrow. Just like, retweet and comment '$BTC'. Random winner in 24 hours. I will announce as a tweet at 12pm EST Sat

I think the best case is to make the "monitor" as un-agentic as possible. We don't mind being "monitored" by various random tiny insects we don't even notice, even guard dogs are fine, but we really do mind if it's a globally connected networked system. And the intermediate case, human security guards with no recording equipment, we mind only a medium amount. The best idea I've come up with so far is to make these monitoring devices out of open-source verifiable hardware that has etched on it an open-weight LLM that classifies for a limited number of activities (basically, just violence; sorry, no scanning for jaywalking or not wearing a mask or drugs or unauthorized manufacturing or illegal immigrants, really just violence) and only records and forwards if that's what it sees. The device has no storage, no ability to load new algorithms, this is all it does. Also, a hash of every recording is published onchain (and a device can't record the next one until it sees confirmation of the previous one) so we can all count how much recording is done. Finally, everyone has a legal right to take any such device they see and grab it and dissassemble and verify the hardware themselves. Explicitly fund university clubs, hacker groups, etc to regularly do this.

I'm giving away 0.1 BTC to one random person! As promised, I'm sending 0.1 BTC (~$6,800) to one lucky follower. To participate, like, retweet, and comment 'Okay'. * You will be disqualified if you are not a follower. The winner will be announced within 24 hours! Grab your chance!

Hello guys, tonight I watched a 50 minute podcast with Albert Castellana from GenLayer on The Future of Crypto Compliance, and one minute in it is still stuck in my head. Code is free now, execution is becoming free, so what still has value? According to Albert, it’s judgment. AI already decides what we see on Twitter, whether we get hired, whether we get a loan or insurance, and right now we’re just trusting a few big AI labs with that power Main points: - His previous project lost access to around $150M after keys were lost, and the courts turned out to be too slow and expensive - GenLayer is building a small digital court onchain for decisions that aren’t just yes or no - Five random validators, each with its own AI model, vote on the outcome - Disagree? Appeal, and the validator set doubles each time, up to 1,500 - The first decision takes minutes and costs around 50 cents - It doesn’t replace real courts, it handles small disputes nobody would take to court anyway

My husband stopped initiating sex after I turned him down three times in one week. At first, I thought he was finally understanding that I couldn’t always be in the mood whenever he was. Then I noticed he had stopped touching me altogether. No random kisses in the kitchen. No pulling me closer when we watched television. No morning hugs before leaving for work. I actually preferred it that way until one evening I asked why he was behaving like a stranger. He looked at me and said, “I figured you wanted me to stop making advances.” I laughed. “There’s a difference between not wanting sex and not wanting you to touch me.” He nodded slowly. “That’s the part I’m struggling to understand. Because lately, every time I reach for you, I have to calculate whether you’re going to feel loved or pressured.”

I am actually disgusted by the timing of this drop. And this will get a bit more technical. Let’s do a real technical teardown of this "1,385-sigma Blackwell weight fingerprint" report that OpenAI and Anthropic are breathlessly promoting today, because banking on nobody reading past the headline is how DC narratives get manufactured. ​First, look at the actual CUDA maths: when PyTorch initialises model weights (torch.nn.init) with a fixed PRNG seed, curand/Philox parallelises random number generation across the GPU’s Streaming Multiprocessors (SMs). Because Hopper has 132 SMs and Blackwell has 148 or 152 SMs, the exact same seed lays out initial weights in a different spatial pattern depending on the GPU’s SM count. In huge MoE models, certain low-gradient null-spaces or unrouted slices preserve faint directional traces of those step-0 weights even after training. ​So yes, brute-forcing the seed and SM grid can show that a model’s step_0 weight initialisation script ran on a 148/152-SM Blackwell GPU. Now read what the summary explicitly admits in the fine print: 1. Initialisation is not full training: The report literally states that these signatures show how the weights were created, and "support (but do not prove) the inference that Blackwell also participated in model training". Initialising a model checkpoint takes seconds on a single rented dev node before loading the safetensors onto a standard Hopper cluster. The "2,000 to 3,000 Blackwell GPUs" and "$500 million" cluster numbers do not come from the weight fingerprint and Depue admits in his own thread that he simply "estimated" cluster sizes from release schedules and model details. ​2. Zero geolocation and zero proof of smuggling: The final sentence of the executive summary explicitly admits that the findings "do not establish the physical location of the chips, how they were acquired, or whether export control regulations were violated". Depue himself rented a cloud GB200 instance just to run his verification test. Commercial labs that are not on the Entity List renting bare-metal Blackwell nodes in offshore clouds (Singapore, Japan, or the Middle East) is completely legal under BIS rules and leaves the exact same SM fingerprint. Now ask yourself why a former OpenAI researcher dropped this on November 12, the exact day Jensen Huang was pushed out, with OpenAI and Anthropic immediately stepping in to "independently verify" it. US closed-source labs are watching their paid API moats get eaten alive by free Chinese open-weights models. Conflating a step-0 PRNG initialisation seed with a "$500M smuggled Blackwell cluster" hands DC hawks the exact pretext OpenAI and Anthropic want to regulate global cloud compute and kneecap open-source AI.

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 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.