A quieter weekend, but three developments worth knowing. Each is explained from scratch, with no assumed background, plus links to go deeper.
Illustration.
1OpenAI shrinks its next funding round
OpenAI is now looking to raise up to $20 billion from investors, down from an earlier target of about $30 billion, CNBC confirmed on Saturday. The round would value the company at $1.25 trillion, still one of the most valuable private companies ever. The deal isn't final, so numbers can move.
Why it matters: a "funding round" is when a company sells a slice of itself to investors for cash, and OpenAI needs enormous amounts of cash to pay for the data centers behind ChatGPT. The trimmed target landed the same week the stock market wobbled over whether all this AI spending will pay off - chip maker Intel fell 25% on Friday after weak forecasts. Nobody is saying the AI boom is over, but investors are starting to ask harder questions about returns instead of writing blank checks.
2DeepSeek releases a prototype AI that manages other AIs
Chinese AI company DeepSeek released an open-source research prototype called dsh-superagent, alongside a design article about building a "super agent." Today's AI agents - software that can carry out multi-step tasks like writing code or preparing reports - usually need a person to assign work, coordinate several agents, and check on them. DeepSeek's prototype watches for requests on its own, turns them into parallel work items, and reports back, so one human can oversee much more work.
Why it matters: DeepSeek became famous earlier this year for building strong models cheaply, and everything it releases gets close attention. The team is explicit that this is research code, not a finished product - it was tested on simulated tasks, not real-world work. But it signals where the race with OpenAI and Anthropic is heading: from chatbots that answer questions toward systems that run whole workflows.
3Claude renames its plans and lifts an annoying limit
Anthropic renamed its Claude subscription tiers over the weekend: the $100-a-month plan is now called Max, and the $200 plan is now Ultra, with prices unchanged and existing subscribers moved automatically. The real change: Ultra subscribers with verified accounts no longer hit a "5-hour limit," a rule that paused heavy use every few hours. The total weekly allowance stays the same - you just aren't forced to spend it in short windows, which matters for long coding sessions that used to stall mid-task.
Why it matters: this one is mostly relevant if you pay for Claude, but it's part of a broader pattern - as AI companies race to win professional users, the practical limits on how much you can use their tools keep loosening. Claude's ladder is now Free, Pro ($17), Max ($100), and Ultra ($200).
Google is retiring Jules, its background coding agent, per its official October 29 notice. New work stops November 11, the tool becomes read-only, and it goes offline December 31. Your code and pull requests already live in GitHub, but export your Jules sessions, settings and memories before the deadline. Google is folding those lessons into its Antigravity platform.
All stories verified against original sources on the morning of November 1, 2026. The funding round, DeepSeek's prototype, and Anthropic's rollout are all still in progress, not finished outcomes.
