Web3 feels flexible. isn't it? AI engineer. LLM
Joined October 2023
- Tweets12.4K
- Following1.5K
- Followers736
- Likes5.3K
Sam Tee retweeted
The crowded trade problem is one of the more counterintuitive risks in markets.
The common assumption is that if a lot of smart people are in the same position, that position is probably correct. The analysis is sound, the thesis is well-constructed, and broad agreement seems like validation. But what crowding actually does is change the exit dynamics entirely.
When everyone is on the same side, the position works until it doesn't, and when it doesn't, the exit is simultaneous. There's nobody to sell to except other holders who are trying to exit for the same reason. The fundamental thesis can be completely right and the position can still produce a painful drawdown purely because the unwind is simultaneous and there's no incremental buyer to absorb it.
The most dangerous trades in crypto are the ones that feel safe because everyone agrees with them. The consensus is often correct on direction and catastrophic on timing, because the consensus getting in is what makes the eventual unwind violent.
Sam Tee retweeted
I’m giving this a 🔥 rating
Sam Tee retweeted
The world runs on opinions.
Who will win elections. 🗳️
Will Bitcoin go up. 🔐
Will a major event happen. ⬆️
Until now, Africans could only argue about it.
With Predera, your opinions can finally have value.
Predict anything. Stake. Win Big!!!
predera.app/
Sam Tee retweeted
New episode of The Information Bottleneck is out 🥳
In this one, we talked with @srush_nlp, a researcher at Cursor and professor at Cornell, about the hottest topic - coding agents! We talked about how Cursor trains Composer, the challenges, and where all this is going.
One point Sasha made stuck with me.
Coding agents work really well, but only when you can specify a clear hill-climbing signal. A problem where the agent knows it's getting better. Karpathy tried this on nanochat a few weeks back, letting an agent run overnight to optimize the validation loss autonomously. The follow-ups were mixed. Sometimes the "improvement" was worse than classical methods.
Most real problems don't come pre-packaged with a reward signal at all. You don't know if a new architecture is better until you've already run the experiment. You don't know if a refactor is cleaner until someone reads it. You don't know if a product decision worked until months later. A lot of what we call hard problems are hard precisely because the signal is missing, noisy, or expensive to get.
I think the next big challenge isn't getting agents to solve problems. It's finding problems you can actually formalize as hill climbing, or building a cheap proxy that correlates with what you care about.
Sam Tee retweeted
Having an opposing party critising an convincing argument really helps to liberate the mind.
It’s the sudden realization that I am in a hole I’ve been digging unconsciously.
- Drafted a blog post
- Used an LLM to meticulously improve the argument over 4 hours.
- Wow, feeling great, it’s so convincing!
- Fun idea let’s ask it to argue the opposite.
- LLM demolishes the entire argument and convinces me that the opposite is in fact true.
- lol
The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.
Sam Tee retweeted
Isaac early access is filling up.
The top Founder Tier member wins an iPhone 17 Pro Max. Top 500 supporters get $10 each.
Invite your friends to qualify; climb the leaderboard and get priority access when we launch 👀
isaac.finance/early-access
Sam Tee retweeted
🎁 Trade With Your WaterX Agent to win $2,500 Airdrop Position Rewards
🔹 Pool 1 $2000: Trade with your agent on WaterX
🔹 Pool 2 $500: Share your position card on X, tag @WaterX_app and hashtag #WaterXAgent
🧵
Sam Tee retweeted
.... @dgrid_ai turns AI from a service into a shared infrastructure of value
No gatekeepers
No hidden logic
Only verifiable computation and aligned incentives.
This isn’t just decentralization it’s ownership of intelligence itself.
Sam Tee retweeted
GM. New week.
Less noise, more signal.
@Dgrid_ai still leaning into aggregation, routing, coordination
while most chase models + compute headlines.
Different layer.
Different pace.
Building where decisions happen.
Let’s see who catches it this wee💪 💪
Sam Tee retweeted
Collective intelligence grows with DGrid AI where distributed contributors provide compute and insights, filtered by Proof of Quality. Dango structures participation, aligning communities to execute proposals and operations smoothly across networks.
Sam Tee retweeted
Endowment-Backed AI Durability:
DeAI infrastructure often collapses when token incentives dry up or hosting subsidies run out—models and data vanish as nodes lose motivation to serve them. @permacastapp leverages Arweave’s storage endowments to create self-sustaining permanence: once data, models, or agent states are uploaded and encoded, upfront one-time payments guarantee long-term accessibility without ongoing rent or governance begging.
This removes the volatility of “pay-per-use” or “community-funded” hosting models that quietly degrade over quarters. Critical intelligence artifacts stay online and queryable across bull and bear markets alike, backed by the protocol’s economic design rather than short-term hype.
It’s the quiet mechanism that lets decentralized AI survive beyond the initial funding cycle.
🤝 Paid partnership
Sam Tee retweeted
The decentralized ecosystem is evolving into a space where intelligence, compute efficiency, and permanent data layers form the core of real progress. Projects like 0G_labs, Dgrid_ai, and @Permaweb_DAO are leading this shift by building the foundational components needed for an intelligent and resilient Web3.
@0G_labs is redefining how decentralized systems handle AI by merging data availability with real-time processing. Instead of treating infrastructure as a passive layer, it creates a dynamic environment where information can be accessed and processed instantly, enabling smarter and more responsive applications.
@dgrid_ai is transforming global compute by coordinating distributed resources into a unified network. This approach unlocks unused hardware around the world and converts it into scalable compute capacity, ensuring that performance grows naturally as participation increases.
@permacastapp provides the permanence needed for information to remain valuable over time. By ensuring that important data is preserved, verifiable, and easily retrievable, it supports the long-term evolution of AI systems, autonomous agents, and knowledge-driven applications.
Together, these projects point toward a future where decentralized systems are not only secure and open but intelligent, coordinated, and built for longevity. The new frontier of Web3 is defined by the seamless interaction between data, compute, and intelligence.
The intelligent era of decentralization is already unfolding.
Sam Tee retweeted
Execution at scale begins with 0G Labs, providing a high performance data backbone that empowers builders to move without friction, while PermawebDAO preserves the outcome, securing knowledge and culture through decentralized governance built for permanence.
Sam Tee retweeted
Sometimes the most interesting innovations in Web3 are not about finance or trading. They’re about how people communicate and share ideas.
When you look closely at Permacast, what stands out is how it reimagines the role of audio in online communities.
Most podcasts today follow a simple pattern: a creator publishes an episode, listeners tune in, and the interaction stops there.
But communities are rarely that passive.
Inside the ecosystem of PermawebDAO, Permacast opens the door to something more community-driven. Audio can become a collaborative layer of discussion rather than just a broadcast.
Picture a scenario like this:
A developer shares insights about a new protocol.
A researcher responds with their own analysis.
Community members contribute perspectives from different parts of the world.
Suddenly, the podcast is no longer just a single voice speaking to an audience. It becomes a shared dialogue shaped by the community itself.
With infrastructure connected to Arweave and the Permaweb, builders can experiment with new ways for conversations to flow across the decentralized web.
And that raises an interesting thought.
What if podcasts weren’t simply shows you listen to?
What if they became open conversations that communities could participate in, expand, and build around?
Platforms like Permacast are starting to explore that possibility — where audio evolves from passive media into a social layer for collective thinking in Web3.
Podcasts have always been one way experiences: the host speaks, the audience listens, and that’s it. But what if the conversation could adapt, respond, and evolve as you interact with it?
That’s the kind of frontier Permacast is exploring. Instead of static audio, Permacast introduces the possibility of AI-powered dynamic episodes.
Imagine this:
• You listen to a discussion, and AI highlights key points in real time.
• Questions you submit could trigger follow-up commentary or clarifications within the same episode.
• Hosts could collaborate with AI to create personalized versions of content for different audiences.
• Communities could generate summaries, translations, or even derivative discussions automatically
Permacast isn’t just a podcast platform — it’s a living conversation network, where AI and humans co-create the media experience.
The bigger question becomes:
Could this be the start of an era where audio isn’t just consumed, but actively engaged with and shaped by listeners themselves?
