MJ retweetledi
MJ
1.1K posts
MJ
@MJbiskey
✨Web 3 Evangelist ♦️| Speaker&Writting✍️ | promoting real value Web 3 projects🤝©
Central African Republic Katılım Eylül 2022
573 Takip Edilen513 Takipçiler
@CEmmanuel00_ @0BL1VIONAI @_Web3newbie @ALHAJIBIGBRO @Baby682002 @bigyardband @CalmlyUnique @chicago_cgv @glowball_io @HyperliquidX Staying active boss
English
@MJbiskey @0BL1VIONAI @_Web3newbie @ALHAJIBIGBRO @Baby682002 @bigyardband @CalmlyUnique @chicago_cgv @glowball_io @HyperliquidX Grand rising MJ
English
@web3righteous Pls engage guy
x.com/i/status/20492…
MJ@MJbiskey
Big face that this Gen feel is wack 😊😊
English
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.
Andrej Karpathy@karpathy
- 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.
English
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.
English
Tested #RealClaw firsthand 💸
1. deposited $29 (25 USDC + SOL for gas),
2. set up an aggressive SOL DCA via Telegram
3. it auto-executed 5 on-chain buys with step-up sizing over 2 days.
>Price impact checks,
>tx confirmations,
> hourly watchdog monitoring
— all from a chat window. The infra behind this is solid.
Nice build @byreal_io 🫡
Byreal@byreal_io
The Agentic Era of Finance has arrived. Meet RealClaw 🦞 Your free AI agent that earns while you sleep. Read on for early access ↓
English
MJ retweetledi
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
English
MJ retweetledi
🎁 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
🧵
English