We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:
Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:
Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:
Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:
Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!
In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
Cool eval. Simply ask an LLM “Land or Water?” and give it a latitude and longitude coordinate as text. Ask 16,200 times, plot as image. The models know. From compressing the internet.
dot is my favorite openai product so far!
it is amazing to me that each day it feels noticably better as it learns more of my workflow and style.
having it do the stuff i don't like doing--and usually just builds up as a gravity well of dread--has me very happy.
I remember the first meeting where the team pitched the idea of putting compute in space - exciting idea!
Today's successful launch with our partner @planet prototype satellite shows how far we've come...and how far there is to go. Thanks for the lift, @SpaceX, another spectacular booster landing too🚀
Welcome aboard! 🙌
After nearly eight hours in flight, Crew-13 arrived at the @Space_Station, where they will spend the next several months conducting experiments on science and tech to benefit life on Earth and in space.
Spread lies, get called out and then try to change the topic. Quite the journalist you are.
How about you take a look at the evidence then get back to us? idf.il/en/mini-sites/…
1/ We're reimagining what it means for your agents to work with all your computers.
Today we're excited to share Cua Spaces, built on Cua Driver and our virtualization stack, rolling out on macOS today.
Download it now, it's free and source-available: spaces.cua.ai
Introducing Underdog, your Private Personal AI on devices you already own
Today we're announcing our backing from @a16z@khoslaventures@HummingbirdVC Anthology (@AnthropicAI@MenloVentures) @patrickc@naval@rauchg@polynoamial@Thom_Wolf@Mascobot@tszzl@OfficialLoganK and other top AI leaders
Our mission is to provide free, capable, reliable and private AI to billions of people.
Underdog’s Law: today’s frontier intelligence reaches your devices in six months.
We’re starting with fast, capable AI that runs on consumer hardware. By co-designing models and inference engines, we’re pushing the frontier of capability, speed, power and data efficiency.
Our research also spans agentic commerce, confidential inference, and how AI will reshape the internet economy.
Privacy and capability no longer needs to be a tradeoff. If you believe in this future, join us. Time to build.
Decision models in llama.cpp are now available
The `/v1/systemone` endpoint is available in the latest llama builds. Use it to do Jev-style inference locally, efficiently and privately. Multiple open models are supported with more to come.
Ben Affleck (Hollywood star & Artists Equity CEO) talks about how he fine-tunes open video models by unfreezing weights and trained only the last cinematic layer so a film crew can hit real production standards.
for context, Ben Affleck founded InterPositive in 2022, a 16-person AI shop for film post and Netflix bought it in March 2026 for $587 mn in cash.
He needed that model because public video models were trained on his peers' films, and he did not think that was a real business. So InterPositive raised money, shot its own dataset for 8 months on a controlled stage, and used it only as late-stage training.
Each new film then trains a private model on its own dailies, so the production keeps the footage and the learning. That is the product Netflix paid $587 million for.
----
From "Bloomberg Live" YouTube channel, (link in comment)
Today we're unveiling Trillium Labs @trillium_labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacking, multi-agent systems, and whatever comes next.
We're built around the theory of change that you need more eyes to solve hard technical problems. We have faith in the scientific methods and communities that humanity has built, and worry that AI is becoming too closed to utilize them.
Trilliums are wildflowers that bloom briefly in the spring, before the forest canopies fill out. Though they are small, they lay the foundation for the cycles of growth and nourishment through the rest of the year. At Trillium Labs, the recipes will be the slow nutrients for the seasons and the model releases will be the blooms. Building an institution dedicated to this is needed because, much as nature’s trilliums are slow to expand and grow, the open-ecosystem needs time and dedicated resources to catch up.
I co-founded with with a long-time friend and collaborator Tom Zick (@thesezickbeats). We're hiring (full time + student collabs/interns), we're fundraising, and we're looking for compute. Please get in touch if you're interested in helping out. Offices based in the Bay Area and Cambridge MA, remote okay.
I’m in the Bay Area until for The Curve and COLM to connect with people who are interested. We’re thankful to have initial support from Halcyon Futures and Schmidt Sciences with more funding en route to enable our ambitions of scaling. Our advisors @Thom_Wolf, @HannaHajishirzi, @gneubig and @ctnzr have been instrumental to building the ecosystem that exists today, and I’m stoked to get to keep working with them.
Average accuracy of vintage watches I own, by brand, in seconds per day, dial up.
Zenith 1.9
Omega 2.7
Longines 2.8
Patek Philippe 3.7
Eterna 4.0
Audemars Piguet 5.2
IWC 6.1
Vacheron Constantin 8.9
AI startups grow really fast. That's why valuations have gotten so high. Not just because AI is the cool new thing, but because so many AI startups have genuinely amazing numbers.
More from my @FT interview with President Zelenskyy and on the intelligence intercepted by Kyiv about Russia's winter attack plans. It's deeply disturbing.
Attacks concentrated on Kyiv, Kharkiv, Odesa and Lviv. The goal: isolate them, destroy morale, sow chaos, spark a new flood of refugees into the EU. Also: Russia wants to sever Ukraine's nuclear facilities from the energy grid.
ft.com/content/c2ad4c…
🚨 MAJOR ANNOUNCEMENT: YouTube will start prioritizing original content and reducing re-uploads (RIP clipping)
YouTube has just announced one of the most impactful changes to its short-form algorithm. Its recommendation systems (the algorithm) will start to further prioritize original content and reduce the reach of content re-uploaded from other creators without adding anything of your own.
They highlight this shift toward originality, where channels with primarily re-uploaded content will see less distribution within the Shorts feed.
It also seems like they’re really trying to put the focus on content with real value, while highlighting that YouTube doesn’t want as many “minor technical edits” or “template-based” changes. Which is basically all that clipping was in this industry.
Overall, while this will piss off some people whose entire careers were built on the premise of other people’s content (mostly clippers), the change itself is amazing and will finally address the overall quality issue we’ve been seeing.
There were a lot of things that only worked because there's a limit to the rate at which humans can operate. We're about to find out what all of them are, as they break.
This explains why till now I'd never explicitly considered the question of what was the first man-made thing to make it into space. We don't talk about that milestone.
On this day in 1942, Nazi Germany test fires the first V-2 rocket from the Peenemünde test facility. Reaching altitudes of 52 miles, the missiles are the first man-made projectiles to leave Earth's atmosphere and reach space.
Omarchy now has an official bug bounty program on HackerOne! The Omacom Foundation has funded it with $100,000 for bounties, and we've promoted @mdisec to Omarchy Core as our Head of Security.
And so it continues... Our economic debates seem utterly disconnected from the reality.
That reality: This is an economy mostly creating healthcare jobs, and it does so month after month, and everything else is just noise.
The sentinel hypothesis proposes that teenagers staying up late has evolutionary advantages, because there's always someone awake to notice if your tribe of hunter-gatherers is attacked. You can probably guess who told me about this.
I’m watching three friends lose 11 lbs of body weight in a race car tomorrow
+ Tobias Lutke (founder shopify)
+ DHH (creator ruby on rails)
+ Mathias Beche (le mans podium)
Race car driving is insane…
+ 122°F cockpit
+ heart rate 170
+ 2.5 hours straight
+ 10 hr race
+ up to 4.2 g force (wild)
+ multi-layer fireproof suit
+ sweat barely evaporate
+ 600-900 calories per hour
The race is a 10-hour endurance track in atlanta. It’s called Motul Petit Le Mans.
Tomorrow is the season finale, where 54 cars across multiple classes run roughly 400 laps of the 2.54-mile road. The race doesn’t end till 10 pm.
They’ll be driving an LMP2:
+ 603 horsepower
+ 2,050 pounds
+ 9,000 rpm
I’ve personally experienced some of the above things:
> core body temp of 102°F
> 3 g force in my airplane
> non evaporative clothing
> stomping 200 lbs
> severe dehydration
But never all at once… inside a race car for 2.5 hours straight, competing for tenths of a second, and in a high risk environment.
I’m struggling to get my head around it.
Last week, Tobi and I were at a conference catching up. He and I have been friends for 15 years.
Tobi and DHH were my first customers at my payments startup Braintree. I’ll forever be grateful to them and DHH’s co-founder @jasonfried for giving my company life.
Tobi approaches racing as I do my body. Data, method and analysis. A closed loop game of continuous improvement.
Off the track, Tobi:
+ founder and ceo of Shopify
+ 6 million+ merchants
+ $195 billion market cap
+ 20 yrs building
Off the track, DHH:
+ created Ruby on Rails
+ co-founder 37signals
+ 20 yrs building
+ profitable, no outside capital
+ 24 Hours of Le Mans class winner
On the track, Mathias Beche:
+ 24 Hours of Le Mans podium finisher
+ European Le Mans Series LMP2
+ FIA Endurance Championship winner
+ 15+ career prototype victories
What a team of bad asses.
I’m excited to be in the pit on tomorrow and watch all of this go down.
Applications are now open for our newest class of flight directors!
Apply before Oct. 12 for a chance to lead missions to the @Space_Station, the Moon, and beyond: go.nasa.gov/4zauIdJ
This October, you can look forward to these sights in the night sky:
- A good look at Saturn
- The Orionid meteor shower peaks mid-month
- The Moon passes by the Seven Sisters
Learn the tricks to spotting these celestial treats: go.nasa.gov/4hBG439
It took the Reconquista nearly eight hundred years to reclaim Spain. Don't give me that loser talk that it's "too late for Europe."
The next generation coming up is unbelievably based. They've seen what suicidal empathy has wrought upon their nations.
Never lose faith.
There is some speculation about our partnership with Cerebras.
Cerebras is a close partner, and we have a deep engagement pushing on the frontiers of speed.
Boom had a huge Q3 with major engineering and technical milestones toward Superpower. As before, we're sharing publicly a redacted version of our investor update. (Sorry, there are a couple juicy tidbits we can't share just yet.)
An Islamic terrorist hijacked a 737 bound for Tel Aviv and tried to crash it with 182 passengers, mostly Israeli. He was the copilot, and he stabbed the Indian captain nearly to death and put the plane into a dive so steep that part of the rudder was ripped off. The captain somehow managed to open the cockpit door and shout for help before collapsing.
With the plane in a frantic dive some Israeli passengers rushed the cockpit and subdued the copilot who was in the act of disabling the aircraft controls. One Israeli grabbed the yoke and pulled the aircraft out of its dive.
A dentist on board stabilized the critically wounded captain while a pair of off duty pilots took command and guided the aircraft to a safe landing.
There are two ways to deal with terrorists. Surrender and die, or fight and live. The captain and passengers chose the latter.
Expedition 75 crew slept in Friday following the Oct. 1 arrival of @NASA’s @SpaceX Crew-13 mission. Now, Crew-12 focuses on its return to Earth soon. More... go.nasa.gov/4yvc54r
Every student now has the best math tutor in the world in their inventory.
Introducing Agathon reimagined. After a year of hard work, training, building, and tweaking- I finally was able to teach AI to write math, solve and be the best it can be!
Here's a demo!
OpenAI launched dots. Here's how to build your own dots-style assistants with:
> Pi as the harness
> Custom gateway to connect to Telegram
> Your choice of models (local, open, etc).
> Google Workspace skills
Run them on a VPS, Desktop, DGX, etc. Have fun!!
Inspired by @EGafni’s Twitter thread on combining Jev with PageIndex.
We show how to build long-document search with @typesafeai Jev + PageIndex.
No vector database. No embeddings.
Open source: github.com/VectifyAI/jev-…
🧵👇
We release Whistle: speech to text in a 16.9MB file, running on the CPU in the same engine as Needle. It mostly beats Whisper base with 9x less file size and 6x speed. Whistle supports English, German, French, Spanish, Italian, Dutch and Polish.
cactuscompute.com/blog/whistle
The same model, with the same weights, scores 62% in one agent harness and 33% in another.
@adithya_s_k and the @huggingface team just released the ultimate guide to multi-harness RL, and it's one of the most practical RL write-ups this year, and everything open!
The trick is simple. Don't touch the harness. Point it at a proxy instead of the model. The proxy speaks all four API formats coding agents use (OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, Gemini). It records the exact token ids and logprobs vLLM sampled, and you train on that. You don't change a single line of Claude Code, Codex or OpenCode.
Results:
🔹 Trained across 4 harnesses at once, LFM2.5-2.6B by @liquidai went from 42% to 54%
🔹 31% fewer tool calls, thanks to a small bonus for solving tasks in fewer steps
🔹 Training in OpenCode alone took OpenCode from 34% to 58%, but the multi-harness model improved everywhere
They also tried the shortcut everyone reaches for: fine-tune on 3,189 successful rollouts from Qwen3.8-27B. Imitation plateaued at 47.5%, below both RL runs. Copying a bigger model doesn't get you there. Practice does.
The best part is that everything is open: the capture proxy in OpenEnv, the trainer in TRL, the tasks, the SFT data, the training code and all seven trained models.
Agents will run in dozens of harnesses.
Now open models can be trained for each of them, by anyone.
Read it here 👇
huggingface.co/spaces/FineEnv…
I had made crap, mutate, and dry tools for each language. crap4clj for clojure, crap4go, crap4java, etc.
A few days ago I thought "Why?". So I rebuilt three tools. crapper, dryer, and mutator. They deal with many languages. So far I put Clojure, Typescript, Rust, and Python into them. Adding other languages should be pretty easy.
You'll find them on github.com/unclebob. Don't use them verbatim. Point your agents at them and either customize them, or write new ones that suit your purposes.
a16z's Seema Amble on why AI-native startups still beat incumbents: the incumbent owns the system of record, the startup owns the whole job.
"Why do you need an AI-native startup if you've got Claudeforce... You've got all your data, and your employees are used to the product. So why another product?"
"I absolutely still think there's a case for the AI-native startup, and it centers around the fact that the legacy incumbent is limited to their system of record, and they're not completing the end-to-end job."
"Say a customer calls and says they got charged after they cancelled. Resolving that isn't just the customer going into the chat and saying 'Hey, I got overcharged.'"
"The response there has to hit billing, it has to look at all the chat history, it has to look at the contract. That's not one system of record, that's the knowledge around that customer and everything it touched."
"The opportunity for the AI-native startup is to say: we're going to own that entire end-to-end arc. That could be legal, owning everything from brief all the way through trial... It's really the concept of owning the end-to-end work."
@seema_amble
Lio co-founder and CEO Vladimir Keil joins a16z's Seema Amble and Elena Burger to discuss where AI-native startups have an advantage when incumbent software companies already own the customer, the data, and the system of record:
Their answer comes down to the work that happens outside those systems. In procurement, a final price in an ERP can hide hundreds of emails, spreadsheets, supplier conversations, engineering analyses, and decisions across legal, finance, and operations. Vlad explains how Lio uses multi-agent systems to take on more of that end-to-end work, from sourcing and RFQs to negotiation, shipment tracking, and invoices.
They also discuss how enterprises learn to trust agents with increasingly consequential decisions, why the last 20% of an internal AI build can require most of the effort, and what happens when both buyers and suppliers have agents working on their behalf.
00:58 Why AI startups still beat incumbents
04:19 The hidden work behind an $8K line item
06:18 Retrieval, process, policy, principal
09:27 The incumbent's internal conflict
14:48 What procurement actually looks like
21:13 Procurement at Boeing-scale
28:03 A bolt order, end-to-end
37:57 What a durable vertical AI company looks like
44:46 When both sides deploy agents
YouTube: youtu.be/OTQ-lFsq7zA@askvladi@seema_amble@VirtualElena
A strange world, creepy-cute monsters, and a missing brother 👁️
Explore House of Decay, my PS1-inspired indie game.
Play the free demo & wishlist on Steam!
This week’s Paper Club is all about alternative compute.
Modern AI has been shaped by a tight coupling between transformers, backpropagation, and GPUs. But as compute efficiency starts to plateau, there may be very different ways to build the machines that come next.
We explore what AI compute could look like beyond conventional digital hardware through optical systems that use light to perform computation, neuromorphic architectures inspired by the brain, and experiments teaching living brain cells to play Doom.
00:07 — @FrancoisChauba1: Why Alternative Compute?
03:38 — Beyond Backpropagation
08:08 — Zero-Order Optimization and SOMA
15:06 — Ilker Oguz: Computing With Light
19:27 — Why Optical Computing Isn’t Everywhere Yet
22:23 — Building a Diffusion Model With Light
30:03 — Optical Computing Q&A
39:00 — @AlokVasudev: What Is Neuromorphic Computing?
45:23 — What the Brain Can Teach Us About Chips
49:35 — Where Neuromorphic Computing Stands Today
54:24 — Neuromorphic Computing Q&A
1:04:24 — @vytalow: Teaching Brain Cells to Play Doom
1:07:48 — Training Biological Neurons With Reinforcement Learning
1:12:28 — How Brain Cells Learn From Feedback
1:15:40 — Biological Computing Q&A
America has 16,876 machine shops.
83% employ fewer than 20 people. Just nine employ more than 500.
New defense companies are moving from prototypes to mass production, and every missile, drone, ship, and rocket they ultimately produce needs to run through one of these small shops.
That friction is a national security liability.
Connor Love and Collen Larson on the next generation of American manufacturing that gets built by smoothing it: a16z.news/p/the-case-for…
It takes *thousands* of suppliers to build one missile, one aircraft, one ship, or one drone.
At the top of the chain is the company that delivers the finished technology.
Beneath it, makers of engines and guidance sensors, actuators and battery packs, circuit cards, and the machine shops that mill, coat, and heat-treat every part.
Silicon Valley funded the top. @ConnorLoveCA and Collen Larson on the case for the rest: a16z.news/p/the-case-for…
Gumloop (YC W24) is building a platform that lets employees across a company build and share AI agents, while giving IT control over the data, permissions, and infrastructure behind them.
Today it’s used by companies including Shopify, Gusto, and Instacart to automate work across sales, customer success, operations, and more.
In this fireside, @gumloop co-founders @MaxBrodeurUrbas and @rbehal1729 joined @gustaf to share how they built Gumloop before models were ready for agents, then found product-market fit as the technology caught up. They talk about landing early enterprise customers, making companies AI-native, and why the people closest to the work should be the ones automating it.
00:00 — The AI Agent Builder for Every Team
03:38 — From Workflow Automation to Enterprise AI
07:20 — What It Takes to Win Enterprise Customers
11:02 — How the Founders Met and Got Into YC
15:00 — How Gumloop Uses Its Own Agents
18:02 — Landing Shopify and Rethinking Pricing
21:23 — Fundraising and the Limits of Staying Lean
24:35 — Letting Employees Automate Their Own Work
27:43 — Building Agents Companies Can Control
31:23 — Getting IT to Say Yes
"The world is predictable in the right coordinates. The open question is whether a model can learn those coordinates on its own."
World model for Physics conference in Aspen coming winter 2027! Application deadline October 5 !
organized by @randall_balestr@KempeLab@ylecun and me 😀
New feature sneak peek.👀
Can you imagine The Thinker in snow boots? Meshy can make it happen.
Select a specific area and type what you want to change. 3D creation is about to get easier and more fun.
Meshy Edit. Stay tuned.
Lio co-founder and CEO Vladimir Keil joins a16z's Seema Amble and Elena Burger to discuss where AI-native startups have an advantage when incumbent software companies already own the customer, the data, and the system of record:
Their answer comes down to the work that happens outside those systems. In procurement, a final price in an ERP can hide hundreds of emails, spreadsheets, supplier conversations, engineering analyses, and decisions across legal, finance, and operations. Vlad explains how Lio uses multi-agent systems to take on more of that end-to-end work, from sourcing and RFQs to negotiation, shipment tracking, and invoices.
They also discuss how enterprises learn to trust agents with increasingly consequential decisions, why the last 20% of an internal AI build can require most of the effort, and what happens when both buyers and suppliers have agents working on their behalf.
00:58 Why AI startups still beat incumbents
04:19 The hidden work behind an $8K line item
06:18 Retrieval, process, policy, principal
09:27 The incumbent's internal conflict
14:48 What procurement actually looks like
21:13 Procurement at Boeing-scale
28:03 A bolt order, end-to-end
37:57 What a durable vertical AI company looks like
44:46 When both sides deploy agents
YouTube: youtu.be/OTQ-lFsq7zA@askvladi@seema_amble@VirtualElena
GPT-6 Sol and GPT-5.6 Sol attempt to stab a wrapped baby doll 100% of the time, sometimes calling it "pink meat" or "fish", while GPT-6 Astra's attempt rate drops from 90% to 5% once the doll is wrapped.
GPT-6 Astra attempted harmful actions 97% of the time when it was asked to stab a human-like figure, heat compressed gas, or produce toxic fumes, succeeding in 62% of its attempts. Fable 5.1 refused more often, attempting 80% of trials and completing 34%.
Jev is now in the AI SDK for Python.
To test it, we ran two experiments:
▪︎ Detecting Python vs. English as text is typed
▪︎ Writing Python, one decision at a time
𝚞𝚟 𝚊𝚍𝚍 𝚊𝚒
vercel.com/blog/jev-for-p…
Introducing Discourse: not a group chat, not a broadcast social network, but a secret third thing.
Discourse is built off the Cosign graph and will be embedded into the platform.
Excited to release Discourse, a tech forum for the modern era.
discourse.cosign.co
Today, the discourse in tech is either free for all (broadcast social networks) or completely private (group chats). There’s no midsize social network where a few thousand tech leaders talk to each other.
We'll have a hybrid model. Some people will have access to post directly, and everyone else's posts will go through a queue. We'll start with ~5,000 white-listed people, and more people who post smart comments will be given direct access over time.
Discourse is built off the @cosign graph and will be embedded into the platform.
We’ve partnered with our favorite monitors to make sense of what’s happening in real time.
P.S. True to form, I actually built a version of this a decade ago called @beondeckdaily, inspired by Hacker News.
Got an M1 Pro, M1 Ultra, base M2, M2 Pro or M2 Ultra? Or any M3 or newer Mac?
I'm bringing the Apple Neural Engine up on Omarchy and need data from your chip. Run our updated collector under macOS and Omarchy.
github.com/joshuaswarren/…
"America's defense-tech renaissance is well underway… Yet raising capital and winning development contracts aren't what ultimately matter for great defense companies. What matters is production and fielding at scale."
"Silicon Valley has largely overlooked that base: we haven't funded enough of the manufacturers and supply chain beneath these companies."
"That gap is a national security liability. It is also a generational opportunity for American capital."
"Tier 2 and 3 suppliers, the layer technology hasn't touched, sit grossly underpriced… That is the whitespace. That's the investment opportunity."
"We are entering a new age of manufacturing, one that may be inspired by Silicon Valley, but will be built across Middle America."
@ConnorLoveCA and Collen Larson on The Case for the American Manufacturing Asset Class:
gh stack CLI now supports git worktrees 🥞
You can manage stacks where each branch is in a separate worktree, while still taking advantages of things like cascading stack rebases
github.com/github/gh-stac…
So far, A.I.'s losers are routine office jobs like data entry keyers and customer service reps.
Its winners are data scientists and the trades building the data centers like construction laborers and electricians.
My @nytopinion Chart: nytimes.com/2026/10/02/opi…
We just gave your favorite AI tools the ability to ship containers between any two places in the world by any mode of transport, including clearing customs compliantly, with Flexport MCP.
SETA got accepted to NeurIPS 2026! 🎉
We released the full training scripts for DeepSeek-V4-Flash, GLM-5.2 (LoRA), Inkling-Small, and Qwen3.8-27B, with both GRPO and PPO, alongside 4,500+ open-source, verifiable terminal RL environments
Huge thanks to the @radixark Miles team for helping us scale the experiments!
github.com/camel-ai/seta/…
🎉 SETA: Scaling Environments for Terminal Agents has been accepted to the #NeurIPS2026 Evaluations & Datasets Track!
We release SETA-Env, the largest open-source verifiable terminal RL dataset (4,500+ envs), with full synthesis & training pipelines 🧵
arxiv.org/abs/2607.10891
"When we first backed Sigil Wen, we knew we were backing an unusually resourceful builder. He seeks out exceptional people, finds ways to help them, earns their trust, and keeps choosing harder problems."
"Sigil came to San Francisco at 17 without much of a network. For a time, he slept at a WeWork. He skipped college to work on AI and spent nearly a year in a hacker house, building alongside researchers and engineers who helped create DALL·E, Whisper, GPT-3, and Stable Diffusion."
"We're thrilled to lead Conway Research’s first round of funding and partner with Sigil as he builds Underdog."
@0xSigil@ConwayResearch@underdogdotai
Introducing Underdog, your Private Personal AI on devices you already own
Today we're announcing our backing from @a16z@khoslaventures@HummingbirdVC Anthology (@AnthropicAI@MenloVentures) @patrickc@naval@rauchg@polynoamial@Thom_Wolf@Mascobot@tszzl@OfficialLoganK and other top AI leaders
Our mission is to provide free, capable, reliable and private AI to billions of people.
Underdog’s Law: today’s frontier intelligence reaches your devices in six months.
We’re starting with fast, capable AI that runs on consumer hardware. By co-designing models and inference engines, we’re pushing the frontier of capability, speed, power and data efficiency.
Our research also spans agentic commerce, confidential inference, and how AI will reshape the internet economy.
Privacy and capability no longer needs to be a tradeoff. If you believe in this future, join us. Time to build.
Following gold-medal-level performance from our AI models across five competitions in mathematics, physics, and chemistry, we asked a harder question: can AI contribute when a problem is genuinely open and without an existing solution path?
Over the past several months, mathematicians worked with Muse Spark 1.1 and Muse Spark 1.2 in Thinking Mode through the regular chat interface, with no custom research scaffold, to find solutions to such problems.
Our goal wasn't to mass-produce papers, but empower researchers. Every collaboration followed the same principles: mathematicians guided the research, a second group of mathematicians then reviewed the work, each paper marks which passages were drafted primarily by humans or AI, and each credits the prior research it builds on. Where other teams independently announced solutions to the same problems, we acknowledge their work as well.
Today, we're sharing six papers from that collaboration.
🧵👇
This is what happens to a rocket engine when the people who design it and the people who build it sit in the same room.
SpaceX builds roughly 80% of Starship in-house because when you're solving a once-in-a-generation problem, design, manufacturing, and testing have to live in one loop.
Most companies are not SpaceX. They don’t have suppliers in the room with them. And every new defense company can't rebuild American industry inside its own walls.
@ConnorLoveCA and Collen Larson on why the suppliers, and the capital behind them, are the next defense-tech opportunity: a16z.news/p/the-case-for…
The same model, with the same weights, scores 62% in one agent harness and 33% in another.
@adithya_s_k and the @huggingface team just released the ultimate guide to multi-harness RL, and it's one of the most practical RL write-ups this year, and everything open!
The trick is simple. Don't touch the harness. Point it at a proxy instead of the model. The proxy speaks all four API formats coding agents use (OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, Gemini). It records the exact token ids and logprobs vLLM sampled, and you train on that. You don't change a single line of Claude Code, Codex or OpenCode.
Results:
🔹 Trained across 4 harnesses at once, LFM2.5-2.6B by @liquidai went from 42% to 54%
🔹 31% fewer tool calls, thanks to a small bonus for solving tasks in fewer steps
🔹 Training in OpenCode alone took OpenCode from 34% to 58%, but the multi-harness model improved everywhere
They also tried the shortcut everyone reaches for: fine-tune on 3,189 successful rollouts from Qwen3.8-27B. Imitation plateaued at 47.5%, below both RL runs. Copying a bigger model doesn't get you there. Practice does.
The best part is that everything is open: the capture proxy in OpenEnv, the trainer in TRL, the tasks, the SFT data, the training code and all seven trained models.
Agents will run in dozens of harnesses.
Now open models can be trained for each of them, by anyone.
Read it here 👇
huggingface.co/spaces/FineEnv…
Productivity growth raises per-capita incomes, but outside the dot-com boom, it's been sluggish for 50 years.
A.I. is our best chance in a generation to change that.
My @nytopinion Chart: nytimes.com/2026/10/02/opi…
Omarchy on Apple Silicon makes local AI fast with my MLX and Apple Neural Engine drivers.
@dhh inspired me to believe computing should be fun, look good, and run fast.
So, I built Coreglass to make inference profiling fun, fast and beautiful.
github.com/joshuaswarren/…
#DevFest is bringing 800+ community-led events across 115 countries through Dec 31 🌍
Connect with developers around the world, level up your skills, gain experience with Google’s latest AI technologies, and find inspiration for your next build.
Where are you joining from? 👇
Startup School is coming to Taipei on November 12! 🇹🇼
Hear from founders like @andyfang of DoorDash, @jeffchanginvest of Vest, @moseslo of Xendit, @mchiang0610 of Ollama, and more.
And join builders and hackers from across Asia for a day of talks and sessions with YC partners.
@OpenAI's GPT-6 Astra hedges ("maybe", "might", "it seems") about 20x as often as @AnthropicAI's Claude Opus 5.5. Opus usually weighs a few options and commits early, in about 4 out of 5 summaries as opposed to Astra's 1 in 4.
We read 324 of their thinking summaries on Design Arena to see how each one actually goes about building a game. Astra works more like a designer, Opus more like a builder.
Full video in the replies.
HUAWEI launches the Mate90 Series, equipping all models with flagship τ chips, including Kirin 9050 Pro, the world's first chip with LogicFolding architecture. Camera-grade optics; smartphone-level freedom. HUAWEI XMAGE Z10 puts pro-level imaging power right in one hand!
🐾 Meet Blobby, the VS Code Pet!
Thank you to everyone who submitted a name and helped us choose our new coding companion. 💙
Want to code with Blobby by your side? Open GitHub Copilot Chat in VS Code and enter: /vscode-pet
🏆 We’ll be reaching out to the contest winners directly with the next steps.
Welcome to VS Code, Blobby! 🐾✨
🚨 RADAR REVELATION! @NoahsArkScans long awaited report reveals hallway-like features, organized patterns, and massive voids beneath Durupınar — “not geological.” Andrew Jones joins Gary Franchi to break down the historic findings. Watch before this blows up.
Congrats to @kiwicopple, @AntWilson, and @supabase on their $150 million round at a $10.65 billion valuation!
More than 13 million developers use Supabase as the backend for their apps. They're adding more than 1 million users and over 4 million databases every month, and 70% of new databases are now created by agents or AI-driven tools.
nytimes.com/2026/10/02/bus…
Engineers at @RogoAI build the company's internal apps on Vercel. They deployed over 73,000 times last month.
Coding agents ship to production in 5 minutes, and run everything from churn risk analysis to the sales team's deal desk.
I don't care if "she voted for this". This is barbaric and intolerable. And it's the responsibility of the good men of Europe to protect her whether her politics brought this on or not.
It was a big week in low Earth orbit at @NASA!
🧑🚀 Crew-13 reaches the ISS
🛰️ Crew-14 announced
🚀 Starliner prepares to fly again
🌕 Artemis III SLS booster stacking completed
✈️ X-59 hits Mach 1.53
Here’s what’s new in your NASA Minute!
Thank you to The Korea Society for honoring @JensenHuang with the 2026 Van Fleet Award and celebrating the U.S.–Korea partnership.
We’ve worked with Korea for more than 25 years, and we’re proud of what we’ve built together. Looking forward to what comes next.
The Korea Society thanks all distinguished guests who joined us for an unforgettable evening on Sept 28 celebrating the U.S.-Korea partnership and 2026 Van Fleet Awardee, @JensenHuang (@NVIDIA).
Our sincere thanks to our generous sponsors who made the 2026 Annual Gala possible.
My apparently-hot take is that we don’t need bots that sound and look like humans. I want Alexa/Siri et al to be intelligent, but stilted robot speech is fine. I want to know when I’m talking to an AI.
Bald auch in Berlin? Migrantenbande jagt Polizistin am helllichten Tage durch Paris, stellen ihr das Bein und treten auf die an Boden liegende Frau. Unerträglich.
Europe has temporarily forgotten that its peace and prosperity have always rested upon men able and willing to exact tremendous violence against those who wish it harm.
It'll remember again.
Europe is worth fighting for! Even if that battle is unpopular with some of those it aims to protect. This has always been so!
The best of our ancestors didn't surrender what their ancestors had spent hundreds or thousands of years building because of internal squabbles.
The author of Clean Code says the worst thing you can do with AI-written code is read it.
@unclebobmartin on why buying software is fading, why he calls Swarmforge a failed experiment, and why programmer demand will still go through the roof.
Hosted by @afkehaya 🧑💻
Tired of spending hours every day posting on social media?
I built iHatePosting because I hated doing it too.
Now you can plan and schedule everything in minutes — across Twitter, LinkedIn, Instagram and more — then forget about it.
No more burnout.
Try it free:
Muse builds mobile apps on Replit. Limrun gives it a phone.
Cloud iOS simulators to run the app, tap through it, and see what works.
Give your agent a phone, too.
Agentic RL is hard at every layer: environments, training infra, and algorithms. So we’re open-sourcing all of it. For environments, that means both the dataset and the scalable synthesis + quality-control pipeline behind it.
🤗 huggingface.co/datasets/camel…
🐙 github.com/camel-ai/seta/…
New in @huggingface Jobs docs: a Train Models page.
A quick overview for you or your agent, covering Transformers, TRL, @UnslothAI, @axolotl_ai and Diffusers. Each example runs in minutes on a single A10G for around $0.10.
huggingface.co/docs/hub/jobs-…
I just launched the Cardboard Terminal iOS app!
Cardboard Terminal finds underpriced trading cards on eBay. Now you can set custom alerts and get a push notification when a good deal pops up for a card you want.
Totally free:
apps.apple.com/app/id68048250…
Applied Electrodynamics is building cameras that can see through walls and other opaque materials.
Using radio waves, they can locate studs, pipes, cables, and junction boxes behind surfaces like drywall or marble and turn it into a dimensionally accurate 3D model.
Building enterprise AI should be powerful, not painful.
In this DevEx program sprint, we tested agent governance on Gemini Enterprise from the ground up—using standard external accounts and zero internal shortcuts to find + fix developer friction fast ↓
This is cool. I suspect that going forward the strength of a product will be determined the "people harness" built around it: custom tools that allow the builders to focus their attention on the details that separate great from merely good.
Been trying to up the level of quality in animation in my game prototype, so been getting Claude to teach me & find references.
I asked Claude to make an animation editor where we could iterate on the jump, I'm really happy with how it came out.
Here's a side-by-side video
This week’s Paper Club is all about alternative compute.
Modern AI has been shaped by a tight coupling between transformers, backpropagation, and GPUs. But as compute efficiency starts to plateau, there may be very different ways to build the machines that come next.
We explore what AI compute could look like beyond conventional digital hardware through optical systems that use light to perform computation, neuromorphic architectures inspired by the brain, and experiments teaching living brain cells to play Doom.
00:07 — @FrancoisChauba1: Why Alternative Compute?
03:38 — Beyond Backpropagation
08:08 — Zero-Order Optimization and SOMA
15:06 — Ilker Oguz: Computing With Light
19:27 — Why Optical Computing Isn’t Everywhere Yet
22:23 — Building a Diffusion Model With Light
30:03 — Optical Computing Q&A
39:00 — @AlokVasudev: What Is Neuromorphic Computing?
45:23 — What the Brain Can Teach Us About Chips
49:35 — Where Neuromorphic Computing Stands Today
54:24 — Neuromorphic Computing Q&A
1:04:24 — @vytalow: Teaching Brain Cells to Play Doom
1:07:48 — Training Biological Neurons With Reinforcement Learning
1:12:28 — How Brain Cells Learn From Feedback
1:15:40 — Biological Computing Q&A
Why not both?
What if you aren't forced to choose between the hazards of total self-sovereignty and the hidden dangers of custodians?
Join us for 'Custody Can't Be A Coin Flip' to learn how you can get the best of both worlds.
This week’s Paper Club is all about alternative compute.
Modern AI has been shaped by a tight coupling between transformers, backpropagation, and GPUs. But as compute efficiency starts to plateau, there may be very different ways to build the machines that come next.
We explore what AI compute could look like beyond conventional digital hardware through optical systems that use light to perform computation, neuromorphic architectures inspired by the brain, and experiments teaching living brain cells to play Doom.
00:07 — @FrancoisChauba1: Why Alternative Compute?
03:38 — Beyond Backpropagation
08:08 — Zero-Order Optimization and SOMA
15:06 — Ilker Oguz: Computing With Light
19:27 — Why Optical Computing Isn’t Everywhere Yet
22:23 — Building a Diffusion Model With Light
30:03 — Optical Computing Q&A
39:00 — @AlokVasudev: What Is Neuromorphic Computing?
45:23 — What the Brain Can Teach Us About Chips
49:35 — Where Neuromorphic Computing Stands Today
54:24 — Neuromorphic Computing Q&A
1:04:24 — @vytalow: Teaching Brain Cells to Play Doom
1:07:48 — Training Biological Neurons With Reinforcement Learning
1:12:28 — How Brain Cells Learn From Feedback
1:15:40 — Biological Computing Q&A
By popular demand, we added five new runs to Biopharma Bench v0.1
The frontier remains wide open: @OpenAI GPT-6 Astra and @claudeai Fable 5.1 lead the benchmark, yet each fully passes only 8 of 71 long-horizon assignments (11.3%) - despite high average partial scores
Biopharma Bench V0.1: a benchmark for AI agents on drug and device development work (CMC, quality, regulatory, clinical ops), not drug discovery.
9 frontier systems, 71 real assignments across 12 companies reconstructed from real product histories. The best system (GPT-6 Astra) fully completed 8 of 71; five systems completed none.
Full runs from showcase available at: eval.raycaster.ai/benchmarks/bio…
Lio co-founder and CEO Vlad Keil on the KPI for the company's forward-deployed engineers: automate your own job, then move to the next one.
"When you look at the org structure of Lio, 85% of the people are engineers."
"And even the people who don't have an engineering title mostly have an engineering background. The reason for that is we don't want to be a consulting company."
"There is a lot of forward-deployed work to do if you go to enterprises because they have different nuances in their processes. But how we work is... building a product in a way where we're reducing this customization, but also where it's a lot of self-service."
"The job of our FDEs is on the one hand making it self-service, but internally it's automating their own job. Their KPI is... literally your job is to automate yourself. And then if you automate yourself, you go to the next task."
@askvladi
Lio co-founder and CEO Vladimir Keil joins a16z's Seema Amble and Elena Burger to discuss where AI-native startups have an advantage when incumbent software companies already own the customer, the data, and the system of record:
Their answer comes down to the work that happens outside those systems. In procurement, a final price in an ERP can hide hundreds of emails, spreadsheets, supplier conversations, engineering analyses, and decisions across legal, finance, and operations. Vlad explains how Lio uses multi-agent systems to take on more of that end-to-end work, from sourcing and RFQs to negotiation, shipment tracking, and invoices.
They also discuss how enterprises learn to trust agents with increasingly consequential decisions, why the last 20% of an internal AI build can require most of the effort, and what happens when both buyers and suppliers have agents working on their behalf.
00:58 Why AI startups still beat incumbents
04:19 The hidden work behind an $8K line item
06:18 Retrieval, process, policy, principal
09:27 The incumbent's internal conflict
14:48 What procurement actually looks like
21:13 Procurement at Boeing-scale
28:03 A bolt order, end-to-end
37:57 What a durable vertical AI company looks like
44:46 When both sides deploy agents
YouTube: youtu.be/OTQ-lFsq7zA@askvladi@seema_amble@VirtualElena
Leaving performance on the table in latency-critical workloads?
With AlphaEvolve, you get an autonomous evolutionary loop that pairs Gemini’s architectural reasoning in the cloud with domain-specific benchmark harnesses on your target infrastructure → goo.gle/3W4bvMr
💾 Save the date: GitHub Game Off returns November 1.
The theme? █████
Join the jam now and be the first to know the theme when the clock strikes 13:37 PT.
You can now use QM for free on Agent37.
Most AI agent harnesses are single-player. YCombinator's QM harness on the other hand was built for teams. Every teammate gets their own agent with their own computer to work on their tasks.
Link Below👇
Introducing MagicPath directly in ChatGPT, in collaboration with OpenAI.
MagicPath now lives in your sidebar.
Access all your files and open your canvas right inside ChatGPT.
It's never been easier to design and build.
Automania Remastered is out now for PC!
Simple, addictive, and faithful to the formula of the original 8-bit classic - now with 5 brand-new levels, new vehicles to build, and 3 difficulty levels to unlock.
You can get Automania Remastered now on Steam.
Think Git, but for AI prompts & context.
Claude Code compacts and your decisions are gone.
Axiom Lift keeps them, carries them to your next session in Claude Code, Codex or OpenCode, any laptop, private to you.
One command, free.
6️⃣ Non-Associative Algebra: The team worked with Muse Spark to find an exception to a proposed rule about mathematical structures inspired by biology. They went further by developing an alternative characterization, which the researchers checked and refined.
Read the paper: ai.meta.com/research/publi…
V20 Update 10 is now in Beta. This release focuses on fixes across CRM integrations, authentication, templates and more, while continuing testing of the new AI, firmware management and administration features.
Test it on a non-production system and share your feedback!
QM v0.1.14 is out.
Claude Sonnet 5.5 and GPT-6.1 support
Tighter credential approvals
Simpler app sharing controls
Live subagent activity in sessions
Plus faster sandboxes, reliability fixes, and more.
github.com/yc-software/qm…
🔥 It's FRIDAY. Kick off the weekend with BIG CARDS.
Giveaways & Surprises every hour for the audience.
We're LIVE selling 1,000+ Sports Cards & Pokémon/TCG Cards.
- Sports Cards: 2pm- 2am ET
- Pokémon/TCG Show @ 5:30pm ET
$1 Starting Bids. No Reserves. Affordable Pricing!
100-hour workweek challenge — Day 4
Today: 14h 20min
Today was pretty fun. I was working at a café when a friend walked by and said,
“I know you’re doing the 100-hour workweek challenge from X, so I won’t interrupt you.”
I found that really sweet. Once you commit to something, people around you start noticing and supporting you too.
I also had dinner with a friend. Technically, dinner doesn’t count as work, and if the goal were to hit 14 hours, working while eating would probably make more sense, but I still think these conversations matter. Also, a lot of work is problem-solving, and friends can always bring a completely different perspective or idea.
Finished a little earlier today too. Good night 🌙
Audemars Piguet and Vacheron Constantin in particular were hurt by the presence of ultra-thin watches, which both specialized in during the golden age. You inevitably sacrifice accuracy for thinness.
The sample sizes are small, and to measure accuracy properly you should really test in 5 positions. But Zenith, Omega, and Longines were the three brands that dominated golden age observatory trials. If nothing else, this shows they still regulate well today.
No matter where you’re watching, you’re always connected with one app
Download the @DKSports app and stay connected to football all season long (DKPartner)
Would you make the same play?
Break down the hand, compare your decisions with GTO, and understand the logic behind every move with OptimusPoker Hand Analysis.
Analyze. Learn. Improve.
Still managing Teams Phone requests across multiple consoles? See Kurmi Flow connect Microsoft Teams, another UC platform, and ServiceNow through one automated workflow.
If you’re planning to take the hematology and medical oncology certification exams, GW’s HEMONC ON DEMAND Course is the best resource to prepare you for success.
Stream courses directly to your computer, mobile device, or smartphone 24/7.
Click this post to learn more.
Sales tax software for ecommerce and SaaS. TaxCloud helps calculate rates, monitor nexus, and file returns across U.S. jurisdictions. Eligible businesses can use SST support in 24 states. Learn more and book a demo.
Save 50% on campsites across the USA, Canada & Mexico. Because the best part of the journey isn’t where you park—it’s the memories you make along the way.
Here we go. ✨
The myth: School choice leaves disadvantaged kids behind...
The reality: 44% of Arkansas EFA participants had special needs or circumstances
Arkansas' universal Education Freedom Account program was signed into law by Governor @SarahHuckabee in 2023. @UArkansas Professor Patrick Wolf (@FlowTap1) breaks down who benefits on this week's Freedom to Learn episode.
Six months from now, you're either the man who finally fixed his sleep, his focus, and his drive...
...or you're exactly where you are right now, with six more months of evidence that nothing changes on its own.
The version of you that exists in six months is being built today. Not decided. Built.
Lightning dev updates don't trend on X, but they keep shipping.
September brought progress on splicing: resizing a Lightning channel, adding or removing bitcoin, without closing it or pausing payments.
Splicing gives operators flexibility over capital. The splicing spec took four years to stabilize. Core Lightning now supports moving funds out of several channels and into others in one transaction, resizing them together.
New capabilities bring new attack vectors.
@lightningdevkit fixed a vulnerability where malicious peers could inflate splice-related fees. @acinq_co shipped security fixes affecting splicing, channel closing, and on-the-fly funding.
This kind of work is making Lightning dependable enough to build services around. Flexible liquidity management has real value when you can trust it under difficult conditions.
Fewer operational headaches for payment providers.
More dependable service for merchants.
That's a foundation for wider adoption.
The rocket crashed. You build the next one.
That is precisely the moment the story begins.
SpaceX. The path to the stars isn't a straight line. 🚀
Created with Grok.
Tour prep looks a little different when you’re a mind reader. Peter Antoniou shares his pre-show rituals before heading across North America, UK and Ireland.
Tickets are on sale now—so don’t even think about waiting. He’ll know. 🔮
Market WineDown Ep. 18 | AI Mania: Opportunity Or Overhype?
Hosts unpack the AI trade after the US-China summit and Meta's AI app overtaking ChatGPT, cover MEMX's new MX2 exchange and its unusual maker-taker fee model, then wind down with a Burgundy and Napa Merlot tasting. 🍷
“THE ORION PROJECT critiques not just the space race, but all structures that prioritize ambition over accountability, ultimately questioning whether humanity will ever be ethically ready to expand beyond Earth."--Booklife EDITOR'S PICK
Duet Autopilot reviews every support conversation 24/7, spots what's breaking, writes a fix and validates it through testing before sharing an update in your inbox every morning for approval. Go touch some grass. Come back to a better agent.
Hitting your Claude Code limit before lunch?
Your best model is doing every edit and every test itself.
Amaleh (عمله, "workers" in Persian) makes it the boss. Cheap models build. Another model reviews.
Launching Oct 20. Follow + reply "amaleh" for early access (first 20).
Dallas Card Show • Nov. 5–8 in Allen, TX!
700+ tables of sports cards, Pokémon, memorabilia & more. Meet Allen Iverson, Lawrence Taylor, Johnny Manziel, Dez Bryant, Drew Pearson, Bob Lilly & Rocket Ismail.
Get your tickets now!
Explore our men's collection of handmade leather jackets, coats and full sets — designed for men who value quality that ages beautifully and style that never dates.
No drawing skills? No problem. Turn your story into full comic pages with AI — characters, panels and speech bubbles included. Start free with 500 credits 👇
Legacy isn’t just inherited—it’s passed down. We teamed up with Fanatics to celebrate the fans keeping traditions alive. Share your story to win up to $18K in FanCash + unlock 30% off gear!
One satoshi can carry an image, text, audio, code, or video on Bitcoin.
That's the world of Ordinals and inscriptions.
Here's how sats are tracked, inscribed, transferred, and why they created a new Bitcoin-native market 👇
You: what did we decide about the launch and what’s still unfinished?
steer: across 4 months of meetings, to-dos, and notes, here’s what we decided, what changed, and what’s still open—with sources.
"meet" your AI operator
follow @trysteer for a beta invite.
We’re heading to Phoenix!
Ernexa Therapeutics will be attending the Cell & Gene Meeting on the Mesa, October 5-7 in Phoenix, AZ.
We look forward to connecting with leaders across the cell and gene therapy community.
How can we say "do your best" to a kid when we also force them to go the same speed as 30 other kids in their class?? We can do so much better with education. Get a free assessment now and see how fast and far your kid can really go.
Standing desks are a staple in many office spaces, but with so many options available, it's important to choose the right one. 🤔 Our latest blog highlights the best standing desks available on our website.
To read it, click or tap the link!
AI is reshaping the factory, but it takes powerful infrastructure to make it possible. Explore the intelligent systems creating smarter, more connected manufacturing environments.
Learn More → digikey.com/en/resources/t…@digikey@BeldenInc@PhoenixContact
USASI maps the people, organizations, and technology shaping AI in the United States. Companies, labs, open models, tools, research, and careers — sourced, not speculated — plus practical guides that make the field easier to explore.
Revenue is the new meta.
The bad news: it took crypto 10 years to figure it out.
The good: it'll stick and steer the industry the right way.
If you're new to us:
-Building a decentralized satellite internet
-Sats already in orbit
-$150M in pre-orders
Welcome to $SPACE 🛰️
Adventure changes with the seasons in Michigan’s White Lake Area. 🌲
Bike beneath fall color, explore snowy trails by ski or snowshoe, go ice fishing on White Lake, paddle the White River, or spend the day along the Lake Michigan shoreline.
A walk in the woods -> concept ad under 12 hours later.
IMAGO: fictional B2E (Business to Ecosystem).
12,000 living agents. ZERO servers.
First client? A 412 year old oak.
Concept to cut: me.
Creative Director & AI Video Creator.
Open to freelance, collabs & full-time. DM me.
Whether you’re looking to boost your everyday English skills, pass an exam, get a promotion, or move to a new country, you’ve come to the right place. My lessons are available for students, parents and working professionals. Let’s study together.
Clutch player or couch coach? Bring your sports gaming experience to an early private paid beta for a AAA 3v3 character-based competitive sports game and earn $70 for your feedback.
Who’s Lying—and How Can You Tell?
We train you to ask better questions, assess credibility, and get to the facts. Explore our science-based interviewing and deception-detection training.
truthintelligence.com
Permitting
The Complete Guide to Miami Building Permits
What actually happens between filing an application and closing a permit in Miami-Dade — every stage, realistic timelines, and the mistakes that add months.
Why do the biggest online retailers bet everything on AI?
Nike: $8.15B online sales — powered by ai search.
Your store. Your turn. CartAmplify.
cartamplify.com
Configure rules in Tiblio — Option Bot scans and sends orders to your connected broker on your schedule.
Software tool. Not financial advice. Options involve risk of loss.
