World AIQ @MITDLab1

x.com/i/article/2064…

JUST IN: SpaceX officially unveils AI1, its first AI satellite, with up to 150,000 watts of compute payload.

BREAKING: OpenAI has confidentially filled for an IPO. We are about to witness the most incredible IPO run in history.

JUST IN: Amazon signs multi-billion-dollar deal with Corning to boost fiber optics production for data centers.

NEW: Google orders more than 3 million AI chips from Intel.

The US job market is experiencing a historic divergence: US information technology employment has declined -11%, since the launch of ChatGPT in November 2022. During this period, the tech sector has shed -332,000 jobs, down to 2.78 million. At the same time, private education and health service employment has risen +13%, to an all-time high. This comes as the sector has added +3.16 million jobs. As a result, technology employment is now below pre-pandemic levels seen between 2017 and 2019. AI remains a key force transforming labor market conditions.

what is agent looping for the last two years we prompted agents one task at a time. that is starting to change instead of asking an agent to build the landing page and then driving every step yourself, you set up a loop that handles discovery, planning, the work, checking, and iterating until the goal is met looping is a setup you build. almost any agent harness can run it, it just depends on how you wire it up at its simplest, looping is one agent working on itself: > researches > drafts > checks the draft against a goal > fixes what is weak > runs that cycle again until the work clears the requirements you are not prompting each step anymore. the agent repeats the cycle for you the bigger version is a fleet looping. you give an orchestrator agent a goal, it breaks the goal into pieces, hands each piece to a specialist agent, and those specialists hand smaller jobs to their own subagents the whole tree keeps looping through discovery, planning, execution, and verification until the goal is met one agent looping is like a person redoing their own draft. a fleet looping is a whole team running a project end-to-end you create a goal, and the system runs the loop until it finishes within the reqs you set open and closed looping: OPEN LOOPING is exploratory. it still has conditions and a goal, but you give the agent or the fleet a wide space to move in. it can try different paths, discover things, build something you did not fully spec out this is the exciting end, it is what Peter and others are doing, and tbh it is where I want to spend more time the catch is cost, an open loop with real room to explore burns an insane amount of tokens. for the 90 percent of people without an unlimited budget it is not runnable yet, and pointed at projects with a loose standard it turns into a slop machine CLOSED LOOPING is bounded. a human designs the end-to-end path first: > clear goal > defined steps > an eval at each step > a point where it stops or hands back to you (and feeds back performance data) the agents still loop, but inside framework you built. it gets better every run because each pass feeds the next, and it runs on a normal budget because the path is tight. for most marketing work, closed is the one that pays off today. > the orchestrator owns the goal > the specialists own the steps > the subagents do the narrow work > an eval gate make sure its not slop

Peter Steinberger 🦞 @steipete

2 days ago

Here’s your monthly reminder that you shouldn’t be prompting coding agents anymore. You should be designing loops that prompt your agents.

2K 1K 19K 7.3M 14K

Good take My guess is - demand for intelligence is near infinite - but 80% of workloads will be running on 99% cheaper models within 12-18 months - 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?) - rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though - this leads me to think the limiting factor will be energy and compute, not better models At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially.

Tommy @Shaughnessy119

a week ago

The most basic way AI could blow up imo. I'm not saying it does but this is the most obvious way I can see it happening - Per seat subscriptions are massively subsidized. The flat fee was priced way below what heavy usage actually costs - For real business use you have to move

161 231 2K 1.6M 2K

AI service firms are commanding 30x multiples right now. Yes, thirty. That's why a16z, Sequoia, and YC are chasing services, not SaaS. Most agencies will see this and reach for the wrong move. They'll keep selling hours, bolt on AI, and cut headcount to pad the margin. But that's playing the small game. Here's why: 00:00 Why Services Beat SaaS 01:13 The $1 Software vs $6 Services Opportunity 02:52 Why Managed Growth Loops Matter 04:49 Agents, Loops, and Human Judgment 06:43 How Single Brain Powers AI Service Businesses 07:22 The Services-as-Software Manifesto 08:41 The New AI-Native Org Chart 10:13 Building Outcome-Based Offers 11:13 Final Thoughts

ericosiu @ericosiu

a month ago

x.com/i/article/2052…

15 39 437 507K 2K

Here’s your monthly reminder that you shouldn’t be prompting coding agents anymore. You should be designing loops that prompt your agents.

The moat is real. Tri Chemical owns the core hafnium high-k patents. Samsung tried ADEKA, lost in court, and now buys through SK Trichem, Tri Chemical's JV. Once qualified in a fab, the precursor is locked in. Switching means yield loss. New facility in Minami Alps.

BREAKING: Scientists have engineered a virus that invades and destroys glioblastoma one of the deadliest brain cancers. It also turns the immune system into a relentless tumor-killing force. This could be such a super historic shift in cancer treatment.

This is huge. For decades, computing advanced by cramming in more transistors. This research points to a different future. Researchers just demonstrated a ZnO-Te semiconductor device that could help redefine how future chips are built. "For a frequency quadrupler, the team achieved the same function using one device instead of four transistors, cutting component count by 75%." "The device exhibits Double Negative Differential Transconductance (D-NDT), allowing a single component to perform functions that normally require multiple devices." "Researchers also demonstrated approximately 4× faster signal processing compared to conventional approaches." Acceleration is everywhere!

This is the best site on the internet to learn harness engineering. Free. Completely. Most AI engineers have never heard the term. walkinglabs.github.io/learn-harness-… Bookmark this site. Then read this setup ↓

Rahul @sairahul1

2 days ago

x.com/i/article/2063…

47 168 1K 1.1M 3K

"Books are the closest thing you’ll ever come to finding cheat codes for real life. You can access the entire learnings of someone else's career in a few hours." —@tobi

Elon Musk @elonmusk

3 days ago

@AJamesMcCarthy Read books, talk to people & iterate rapidly with hardware & software

726 2K 17K 1.5M 4K

9 books that'll make you smarter than 99% of people: 1) The Demon-Haunted World by Carl Sagan

The most comprehensive Hermes Desktop tutorial on the internet NOW is LIVE. You'll learn sessions, profiles, artifacts, cost savings, and real use cases for making money and building startups with Hermes agents. Whether you're already running Hermes or haven't started yet, this is the episode for you. @AlexFinn says this is the moment Hermes overtakes OpenClaw. S/o to Alex for walking me through it. "It's now the best way to use AI agents on your computer" I do think the desktop app of Hermes looks almost like an Apple product. Everything you need to know about Hermes Desktop App/agents in 43 minutes This episode is 100% free. No ads. @startupideaspod I just want to see you win on the internet. And I think Hermes can help. Plus, It's fun thing to play with this weekend. Share this with a friend. Link below. YT: youtube.com/watch?v=EJm8Ka… Watch

you're telling me anthropic & google are paying spacex ~$26b a year for compute?!! this is more than half the run rate of openai & anthropic just from compute deals & that doesn't even factor in the rocket launches at all. elon accidentally ended up owning a significant portion of three of the scarcest assets in ai.. power, chips, & physical deployment capability. the best lesson here is that if you’re selling picks & shovels during a gold rush, you don’t necessarily need to find the gold. you just need everyone else to keep digging. & also non software elon is pretty much unstoppable, like prime michael jordan type thing.

The product is the mission.

This is actually wild. Someone just fit 31GB of AI memory into 4GB, running on a MacBook. No server, no cloud bill. It's called TurboVec. Google Research published the algorithm (TurboQuant, ICLR 2026), the kind of paper that usually sits untouched for years. Then one independent dev, Ryan Codrai, just built it. Wrapped the math in Rust and shipped it open source. Now 10 million embeddings that needed a 31GB server fit in 4GB on a laptop, and search faster than FAISS. 16x less memory. Zero training step. 100% local. MIT licensed. The gap between "Google publishes a frontier algorithm" and "one person ships it as a tool you can install today" used to be years. It just became months. You're not waiting on big labs to productize research anymore. You're running it yourself. The people building on research like this the week it drops are going to look at the rest of us in 18 months the way founders look at people who refused to use the internet in 2001. Repo: github.com/RyanCodrai/tur…