@mopan_ai @dill_xyz_ | prev secure3, goldman, yale
Joined April 2022
- Tweets705
- Following1.1K
- Followers1.4K
- Likes1.3K
0xted retweeted
K3.1 Pro works directly in the tools used by the 3D modeling and game industries - Blender, Unreal Engine, Godot, and CAD.
On BenchCAD, which tests 3D reconstruction from multi-view renderings through CAD code generation, it scores 92.7% with tools, approaching Claude Opus 5.5 (96.2%) and GPT-6 Astra (95.9%).
K3.1 Pro can deliver a fairly complete project, from asset construction to scene assembly, animation, and rendering.
0xted retweeted
The recent security incidents involving Bybit, Bitget, and now Binance should serve as a wake-up call for every leader in the crypto industry.
Security should never be an area of commercial competition. It is a shared responsibility to protect our customers and the entire industry.
Over the past decade, major crypto exchanges have made significant progress in protecting private keys. We have seen relatively few incidents involving direct private key compromises in recent years. So far, we have not seen conclusive evidence that Binance’s hot wallet private key was leaked. The greater concern may lie in vulnerabilities in access controls and wallet management systems.
Companies that have never experienced a major security incident should not assume they are inherently safer or immune to attacks.
At OKX, we have built multiple layers of sophisticated security protection and maintained a strong security track record over the past decade. But we also recognize that no system is absolutely secure. We must remain humble, vigilant, and constantly challenge our own security assumptions.
A major security breach at any exchange undermines trust in the entire crypto industry, ultimately hurting every company and every customer.
I believe CEOs across the industry need to come together, just as we did during the Bitcoin fork crisis in 2017, to establish stronger industry-wide security standards and coordinated response mechanisms.
For example, transfers and settlements between verified exchanges could be processed in real time, while large withdrawals to unknown or unverified addresses, as well as transactions granting smart contract permissions, should be subject to additional verification and appropriate security delays.
Crucially, these safeguards should be enforced at the cryptographic signing layer, rather than relying solely on application-level controls. Even if an attacker compromises an exchange’s wallet management system, they should not be able to bypass these fundamental security restrictions.
We should also establish industry-wide mechanisms for real-time security incident notifications, coordinated monitoring of suspicious funds, and rapid responses to potential attacks. Exchanges, stablecoin issuers, and other major industry participants should work together to enable the rapid, lawful temporary freezing of suspected stolen assets when necessary.
We should also work together to accelerate the industry’s transition to quantum-resistant cryptography and strengthen our defenses against emerging AI-powered security threats. These are long-term challenges that no single company can solve alone, and they require coordinated research, shared standards, and collective action across the industry.
Security is everyone’s responsibility. No company can protect the future of crypto alone.
glad to be contributing to this, nice work gary
The 9th covenant of AI Agents: ERC-8468
Agent Safety Framework draft
Addressing the growing number of agent safety issues of late, including payment security, information security, intent security, and more.
ethereum-magicians.org/t/erc…
先舆论压低市场 然后400亿抄底买卡？
Pretty much.
As of the past few weeks, debt markets have almost entirely shut off access to money for any company that doesn’t have substantial revenue independent of AI data centers.
SpaceX closed $40B of debt for GPUs, but we’re investment grade and have Starlink & launch revenue.
That means the 200+ neoclouds are going to see guppy-level death and consolidation. Some will make it through, but not many.
0xted retweeted
On the foundation of the series of base-layer frameworks now complete, the superstructure of AIFi's economic units and economic value was also built today, receiving ERC-8454.
AI economic value can be abstracted into a variety of structural elements — value sources, sinks, resistors, capacitors, masses, conductors, valves, controllers, and more. The financial circuit theory of 2022 has finally been realized in code, with the help of AI.
ethereum-magicians.org/t/erc…
0xted retweeted
看了一下DeepInfer的数。我觉得吧，前几天大家转载的那个SiliconData Index的价格，未来一年估计还得不回头地跌个95%+，具体什么时候影响到GPU Rental rate蛮难预测的。
这个价格体系下，梁圣买张卡估计100天就回本了。一旦有朝一日token价格真的暴跌到影响rental rate，所有股价肯定已经跌了一大轮。
给大家几个数，所有的数都是$ / 1 mm token，基本上1mm token 95% cached = 100-150页的薄书.
- Dylan一直在说Jane Street的变现价是$100-200 USD / 1 mm token
- Opus 5.5 / GPT Astra 的价格在deepinfer的体系下大概是1-2 USD / 1 mm token，比标价便宜个80-95%
- 变态的来了，Deepseek 4.1 Flash off-peak的价格在这是1.5美分，又比SOTA便宜个90%。
- 按照一般的大B300 Cluster，根据deepinfer的计算，一个GPU大概能每小时产出2.7 Bn token, 成本换算过来大概是0.25美分 / 1 mm token，这已经比我想象中便宜很多很多了。之前算过我记得是15美分 / 1 mm token，不过没有计算cache。这个level的成本大概率是大机房+高使用率+超高存储时间，一般机房估计都做不到这个theoratical.
- 这也就是大家看到的80-85%毛利 (1.5 / 0.25)。DeepSeek要是以比这个利润率稍好的功率赚钱的话，一个B300大概2-3个月回本，而不是8个月。真是无与伦比的暴利，比贩毒爽多了，自然是砸锅卖铁也要走私B300。
- 如果B300按照break even算的话，那么成本大概是0.13美分 / 1 mm token. 然后原材料其实是B300 ～600k价格的6%不到（按照产业链毛利计算）。
- 所以最后算下来马斯克如果去完全积压的话，每1mm token的最终生产成本可能是0.01美分, or 0.0001 USD。这基本上跟免费差不了太多了。
这种屠夫版的价格我不知道你怎么看，但是when you see the headline API price of Opus @ 8/40 USD input/output，我只能感觉说只要OSS + Post-train和中国模型蒸馏大法能够持续跟随 + 没有新的breakthrough，token 价格未来一段时间应该是以断崖式的下跌。至于什么时候触碰到rental rate, maybe someone smart can tell me.
finishing up the public repo, so anyone can contribute bench trajectory
github.com/mopan-ai/ai-game-…
so i built aigamebench.com/
an independent benchmark for AI models & agents in games, from game creation and self-play to real player voting.
be at top 0.001% ~ 100k
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
everyone needs to get @vibe42ai by @jiacheng_ai , i'm sipping my margarita while letting my agents cook at home
only in x 😘
Dear Prime Minister Ambani,
Please accept my humble apologies for not realizing that you are the real boss of India.
Naturally, you would prefer to maintain your monopolistic exploitation of the great people of India, but would you nonetheless consider allowing Starlink to compete?
There are many areas of India, as there are in all countries, with no Internet access, which denies children the opportunity for education and for small businesses the opportunity to sell their goods to a global market. Starlink would change their lives for the better.
Thank you 🙏
the quantum risk for "average person" is oversold and exaggerated
over the next 5yrs, the probability of a us debt crisis and the probability of any individual tech company getting existentially disrupted by ai are materially higher than the probability of ai/quantum breaking crypto
so sell ur btc, but sell it into what? cash and bonds that are going to 0? or tech companies facing existing threats while also in a giant reflexive bubble? or yet something else that u did 2hrs of “due diligence” and convinced urself u understand?
the point is, if u r spooked by the ai/quantum risk and sold ur coins, u r almost certainly misjudging magnitude of this risk vs the risks of other assets u r sold ur coins into.
0xted retweeted
JUST IN: Tom Lee says BitMine will stop buying more $ETH when it accumulates 5% of the total Ethereum supply.