For business owners
AI stopped demoing and started working
November 6, 2026 in summary: Thinking Machines released a 594B-parameter model built for long-horizon coding, OpenAI launched a security model that found 81 previously unknown Firefox flaws, and Agility Robotics became the first pure-play humanoid robot company on the stock market. The capability is moving into real work, and the question now is who gets access to it.
- Daily brief
- AI news
- Automation
November 6 was the day AI stopped looking like a demo and started looking like work. A new open model arrived that is built for long, multi-step coding and knowledge tasks. OpenAI launched a model whose job is authorized security research, and it showed its results in real flaws found in real software. And a company that builds humanoid robots started trading on the Nasdaq, with its machines already working in warehouses. Three different announcements, one direction: the capability is moving out of the lab and into jobs that used to need a person, and the interesting question is no longer what the models can do. It is who gets to use them.
Capability is leaving the lab
The clearest sign is how the new models are described. Thinking Machines released Inkling 2 Small, a 594B-parameter model built for long-horizon coding and knowledge work, scoring 31 on the Artificial Analysis Intelligence Index, with open weights planned for next week under Apache 2.0, according to Artificial Analysis. "Long-horizon" is the phrase that matters. The pitch is not a cleverer chatbot. It is a model that can stay on a hard task for hours, which is what actual coding and knowledge work looks like. And because the weights are being released openly, that capability will be available to anyone with the hardware to run it, not just to the customers of one company.
The same shift is happening in the physical world. Agility Robotics started trading on Nasdaq as the first pure-play humanoid robot company on the public market, raising $465 million through a SPAC deal, with its Digit robots already working in warehouses, according to TechCrunch. A robot company listing on a stock exchange is a financial event, but the detail that counts for a business owner is the warehouse floor. These machines are not prototypes at a trade show. They are moving boxes in real facilities, which means the economics of physical automation are now something investors can price and competitors can plan around.
The same power, two very different doors
OpenAI's launch shows the other side of that spread. GPT-6 Cyber is a model built for authorized vulnerability research: it found 81 previously unknown high-severity flaws in Firefox and built working exploits for 948 Linux kernel bugs, according to MarketScale. Access is limited to approved security teams for now. That is the deliberate contrast with Inkling 2 Small. One lab is opening its model to everyone, while another is keeping its most powerful security tool behind an approval process, because a model that can find flaws in the world's most widely used software can be used to fix them or to exploit them.
That split is the story to watch. The capability itself is no longer scarce. What is scarce, and what will decide who benefits, is access. Open weights put long-horizon coding ability in the hands of any small team willing to run it. Gated access keeps the sharpest security tools with vetted teams. Public listings put robot labour on the same footing as any other capital investment. The technology is spreading in every direction at once, and the gatekeeping is happening at the same time.
What this means for a small business
The practical lesson is to stop asking whether AI can do the work and start asking who can run it for you. If your team writes code or handles knowledge work, an open model like Inkling 2 Small is worth a look once the weights land, because it removes the subscription and the vendor lock-in from the equation. If you buy software, ask your vendors whether their security review uses models like GPT-6 Cyber, because automated flaw hunting now has a published track record and a vendor who uses it should be able to say so. And if your business touches warehousing or logistics, watch what Agility's listing does to the price of physical automation, because a public company with $465 million in fresh capital is going to push its robots into more facilities, not fewer. The work itself is changing. The businesses that keep up will be the ones that treat these tools as something to try, not something to read about.