AI & robotics news brief ·
OpenAI reportedly raises $11B at $1.25T valuation
Good morning. OpenAI has reportedly raised another $11 billion at a $1.25 trillion valuation, making it the clear center of today’s AI news. Also in the mix: a warning about hacking from free Chinese models, a piece on the AI bubble, and a fresh batch of tools and robotics work.
Today in AI
OpenAI Secures $11B in Initial Funding at $1.25T Valuation
Why it matters: The funding reinforces OpenAI's leading position in AI development while reflecting investor caution amid a cooling market for AI valuations.
9 Free AI Demo Tools Worth Trying in 2026
Why it matters: Understanding available AI demo tools can help businesses enhance customer engagement and reduce lost sales due to lack of immediate product demonstrations.
OpenAI reportedly raises an additional $11 billion at a $1.25 trillion valuation
Why it matters: OpenAI's significant funding boost highlights the ongoing investment interest in AI amidst a turbulent market, impacting future AI development and regulation.
Please stop asking me when the AI bubble will pop
Why it matters: The potential downturn in the AI market could impact funding sources and investment strategies for major tech companies, influencing the financial landscape.
Free Chinese AI Models Are Getting Scary Good at Hacking
Why it matters: The growing capabilities of open-weight AI models in hacking threaten to change the cybersecurity landscape, making advanced hacking tools more accessible to both attackers and defenders.
Robotics
Ai2 Releases ToyBlocks to Test How Robots Learn and Fail
Why it matters: ToyBlocks enables researchers to explore and address the challenges of robot learning failures, which could significantly advance the efficacy of robotic systems in real-world applications.
Physicl raises $10M to make the digital world real enough for robots
Why it matters: This funding could enhance the capability of robots to operate effectively in complex environments, addressing key limitations faced in the training of AI models.
Agent uses to try
Coding & DevOps · Dots
Turn P2 bug reports into draft PRs with demos
A Dot monitors bug reports, tests a fix in the cloud, and opens a draft pull request with a short video for review. The setup was described as the strongest example OpenAI showed, but no performance numbers were shared. The live demo Dot failed to answer on stage, and OpenAI said Dots can still make mistakes even when they follow their instructions.
Try this prompt
Connect to [GitHub repo] and the [Slack channel]. Watch for bug reports marked [priority level]. When you find one that looks actionable, inspect the issue, propose a fix, test it in your cloud environment, and create a draft pull request with a short demo video showing the change working. Post the PR link in the thread, then stop and ask for approval before you do anything that would send, merge, delete, or submit changes. Tell me what you checked and what still needs review.
Money & Bills · Claude Code
Create a portfolio research workflow with live inputs
They used Claude Code to assemble a private investing assistant that mixes live data, macro signals, and news into one review. The project was meant to produce analysis that went beyond generic stock advice. The thread did not include any verified returns or trading outcomes, but it did say that remembering earlier discussions noticeably improved the quality of the output.
Try this prompt
Set up a small investment research assistant in [this folder or project]. Gather live market prices for [tickers or funds], macro data such as [interest rates, inflation, unemployment], and recent headlines, then produce a dated summary of what those inputs could mean for my holdings in [brief description]. Include your sources, and stop to ask for my approval before any trade, submission, deletion, or other irreversible action. I want a one-command report I can rerun whenever I need an update.
Work & Operations · Claude, ChatGPT
Use Claude to screen real estate deals and tenants
Investors used Claude and ChatGPT to pre-screen deals, compare comps, and check tenant and rent details before doing manual underwriting. People said it saved hours and made early screening much faster, often getting them most of the way to a rough yes-or-no. The downside was that the models could sound certain while being wrong, so the thread treated them as a filter, not the final underwriting step.
Try this prompt
I’m reviewing [number] incoming real estate deals. For each one, read the deal sheet and any comps, rent roll, or market notes I paste in. Pull out the key facts, compare them against [my screening rules], and sort the deals into pass, maybe, or reject with a short reason. Also list any tenant or local rent questions I should verify next. Stop and ask for my approval before you do anything that would be irreversible, like sending the analysis to someone, deleting any input, or submitting a decision.