
Founder & CEO @cambriannetwork. Prev. cryptography @SandiaLabs, CS/RL PhD @ucsantabarbara
NYC / Santa Fe
Joined June 2009
- Tweets3K
- Following1.6K
- Followers4.5K
- Likes53.6K
Pinned Tweet
UPDATED: The Agentic Finance Landscape for Q3 2026. Is your favorite project listed?
Agentic finance now makes headlines weekly. The term is gaining popularity among major financial institutions and global governing bodies. It’s wild to see this reality after launching the AgentFi Report series last year.
Our team painstakingly researched, tested, and filtered hundreds of agents in the last few months. Most agents did not make it into our latest report. We name only the most legitimate projects, along with emerging trends, so you can prepare yourself accordingly for the future of finance.
Check out the latest cohort of the AgentFi Landscape 🧵
FYI: Over the last 3 years, hot tech stocks usually peak in their first month of trading at the end of day 4
BREAKING: Anthropic, $ANT, officially begins trading on the Nasdaq and surges to $115/share, up +39% from its $83 IPO price.
This officially pushes the company's fully diluted valuation above $2 trillion, making it the 6th most valuable publicly traded company in the world on this basis.
As your brain/attention span turns to mush, model capabilities are accelerating. The real lesson from this chart is that if you aren't a master of your domain, you won't be able to write prompts that push the models to their limits
A slide from @AnthropicAI's IPO roadshow. Hits hard and is very true.
AI is soooo much better than how most people are using it.
3
16
515
Increasingly likely
Over the next year, major TradFi security breaches could lead to even faster institutional adoption of blockchains.
Blockchain node software is open-source and has been hardened by a decade of attacks from state actors, and blockchain devs already operate like the world is out to get them. TradFi infra is closed-source, so who knows what flaws are hidden and what black hat agents will find and exploit soon.
The recent widespread software/infra outages may not be random. They could be the first wave of an AI cold war. @iamjoshgreen made an excellent documentary digging into Travis's ideas
There will be no end to the wonders delivered by reinforcement learning
Researchers have used reinforcement learning to teach a 5-tonne John Deere skid-steer loader to perform complex acrobatic manoeuvres autonomously. Trained in simulation and transferred to a real machine, the system can balance on two wheels, navigate rough terrain, spin, and control its bucket before landing safely.
The research demonstrates the potential of AI-driven automation for heavy construction and industrial machinery.
Video Credit: @leggedrobotics / @ETH
Paper: wevlv.co/4kltYys
#engineering #technology
Agentic trading and @CambrianNetwork mentioned! Can't wait to share some things that we're building for this segment of agentic finance
New @cbventures market map alert! 🚨
AI agents are moving from talking about markets to trading in them. But as agents get easier to build, where will lasting value accrue?
My new post maps the agentic trading stack through that lens. Here’s where I’m focused:
• Agentic trading harnesses and terminals that turn a user’s goal into an ongoing strategy, with real retention.
• Market reasoning models and data assets that give trading agents a real edge.
• Identity, audit trails, and performance verification infra built for non-human, financial decision makers.
I care less about shiny demos and more about products that get harder to copy as they grow.
If you’re building here, I’d love to hear from you.
We're now at the stage where AI works well for short movies like this. I believe it took many retries for each scene, and you see single shots from one angle because consistency is still hard to achieve. That will be solved in the next year, and then movies will become as good as the storyteller.
Then we'll be on to harder tasks. Here's a video where I share the pattern we'll see with AI forever:
* current models don't work for a particular task
* new model works about half the time
* new model works most of the time
* Almost no change in task performance with later models
(repeat for harder and harder tasks)
Check out our fresh new logo
New logo. Still the #1 source of onchain & offchain financial data on the market.
Say gmonstera to the new Cambrian aesthetic 🪴
I believe this will be true for two reasons:
1) LLMs already contain a basic understanding of every medical, biological, chemical, physical, and mathematical concept known to man. And they can find connections between the fields
2) LLMs will at least double in capability every year for the next 10 years. So in 10 years, LLMs will be at least 2^10 = 1024 times more powerful than they are today. (I think 3^10 = 59,049 is more likely)
SITUATION DETECTED: Anthropic CEO Dario Amodei says AI breakthroughs will turn cancer, Alzheimer's, heart disease, and mental illness into "relics of the past" in less than a decade, ushering in "an age of miracles."
5
21
1,044
The frontier is accelerating, are you ready? Watch this 1962 interview with Warren McCulloch. He invented the artificial neuron in 1943. I was unnerved to see his cold prediction and acceptance of superintelligence. It's probably best for our mental health that we adopt that mindset
I've summarized where things stand at the AI frontier:
- Unreleased models (Anthropic's Models 3 and 4) are now half-way between AI2027's Agents-1 and 2 (graphs fixed for the bug flagged by the AI2027 authors). @DKokotajlo please let me know if I missed something.
- We're right on track with AI2027's prediction of an automated coder (AC) for April 2027, wild. Public models are now massively lagging unreleased models (1 day vs. 1 week of autonomous high-quality work). Fuelled by RSI, the gap is huge and growing.
- This trajectory, if continued, means a world-changing intelligence explosion in 2027, so the public needs to know. I highly recommend reading more on the topic to grasp what's coming.
I read an article about this in FTWeekend. It said Buddhist cultures believe that all aspects of nature have a spirit, even inanimate things like rocks and trees, and that everything is interconnected. Japanese call this kami, or spiritual essence. So if a computer is a thinking rock, it must have kami too. Westerners are individualistic and hate anything threatening our species' uniqueness in the universe. Also, developing countries see AI as a way to leap ahead, while America is comfortable, so we don't want change
In the bad old days, training neural networks took a lot of work. But now frontier LLMs can directly create or edit neural network weights - this opens the new field of "weight-space programming". Imagine LLMs waving a wand and dropping small, real-time AIs into games or robots or directly modifying the behavior of other LLMs.
Weight-space programming was introduced in an MIT blog published last week. I suspected it worked because their LLM had memorized the blog examples in its training data. So this weekend, I hacked with @mvyletel_jr to vibe code a new game (not in OpenAI's or Anthropic's training data) to see whether Astra or Opus 5.5 could create neural networks to fight. It worked. Here's the game being played by two LLM-generated neural networks. Details below
agentic finance mentioned
2026 has been the breakout year for general personal agents. As general agents move into commerce, Franklin Crypto Managing Principal Julian Love @LastNameLove examines what this shift could mean for AI users, developers and the blockchain ecosystem.











































