The World’s Biggest Law Firm Just Bet $500 Million on AI — and Wall Street Is Watching
AI is moving from the lab into the places that actually run the economy: law firms, factory floors, and investment funds. Today’s stories show what that shift looks like in practice — not in theory. Whether you’re a manager, an investor, or just someone trying to get more out of ChatGPT, what’s happening right now has direct consequences for how work gets done.
A Law Firm and a Data Company Are Quietly Reshaping How Investment Funds Get Built
Fund formation — the legal and administrative process of setting up a new investment fund — is one of the most document-heavy, detail-intensive tasks in finance. Kirkland & Ellis, the highest-grossing law firm in the world, has partnered with Palantir Technologies to build a custom AI tool specifically designed to handle it. The deal is part of Kirkland’s $500 million commitment to AI infrastructure, reported by Bloomberg Law.
Think of fund formation like assembling a very expensive piece of furniture — except every piece is a legal document, every instruction is a regulation, and the cost of getting it wrong is measured in millions of dollars. Lawyers currently spend enormous amounts of time drafting, reviewing, and cross-referencing those documents. The Kirkland-Palantir platform aims to automate large chunks of that process so attorneys can focus on judgment calls rather than paperwork.
For everyday people, this matters because it affects how quickly new investment funds — including the ones that back startups, infrastructure projects, and real estate — can actually launch. Faster fund formation means capital gets deployed sooner. It also signals that big-money legal work, long considered untouchable by automation, is changing fast.
Why this matters: When the world’s most powerful law firm spends half a billion dollars on AI, the legal industry follows. Junior attorneys and back-office staff at firms everywhere should pay attention.
“Kirkland & Ellis launched a multi-year partnership with Palantir to develop a custom AI tool for fund formation.”
WIRED’s 28-Tip Prompt Guide Tells You Something Bigger Than Technique
Prompting — the practice of writing instructions to get better responses from an AI — used to be a niche skill. Now WIRED is publishing 28-tip guides for a general audience, which says everything about where we are in the adoption curve. The piece covers techniques like asking the AI to explain its reasoning, specifying the format you want, and giving the model a role to play before asking your question.
Prompting well is less about magic words and more about being a precise communicator. When you tell ChatGPT “write me a summary,” you’re giving it enormous latitude to guess what you want. When you say “write a three-sentence summary for a non-technical hiring manager who needs to decide in 30 seconds,” you’re giving it a real target to hit. The difference in output quality is usually immediate.
If you’ve ever felt like ChatGPT gave you something generic or unhelpful, there’s a good chance the instructions you gave it were either too vague or missing key context. This guide is aimed squarely at people who use these tools daily but haven’t spent time learning why some prompts work and others don’t.
Why this matters: AI tools are only as useful as the instructions you give them. Getting better at prompting is now a practical workplace skill, not a technical hobby.
“28 tips to take your ChatGPT prompts to the next level.”
3M Built an AI That Answers the Technical Questions Engineers Used to Wait Days For
3M — the company behind everything from Post-it notes to medical adhesives — launched a tool called Ask 3M, an AI assistant designed to help industrial customers solve material problems faster. According to the 3M News Center, the tool gives manufacturers and engineers self-service access to the kind of technical guidance that previously required calling a specialist and waiting for a callback.
3M makes over 60,000 products. Its technical knowledge — which materials work under high heat, which adhesives bond to which surfaces, how to solve a specific sealing problem — lives in an enormous library of documentation that most customers can’t efficiently search. Ask 3M is trained on that library and can answer specific product questions in plain language, in real time.
For a factory manager trying to solve a production problem before a deadline, that difference is significant. Instead of waiting 24 to 48 hours for an expert callback, they can get a qualified answer in minutes. It’s a small shift in one company, but it points toward a broader pattern: specialized institutional knowledge is becoming something you can query like a search engine.
Why this matters: Industrial AI tools like this one could meaningfully speed up manufacturing decisions and reduce costly downtime for businesses that depend on materials expertise.
Also Happening in AI
Tool calling — the ability of an AI to trigger external functions like searching the web or running calculations — got a clear explainer from Towards Data Science this week, which is useful context as AI agents become more common in workplaces. On the infrastructure side, LangChain pushed a security-focused update to its core library on GitHub. The geopolitical side of AI got more complicated: TechCrunch reported that the Trump administration issued an export control order requiring Anthropic to restrict certain AI models, raising questions about which competitors stand to benefit. Meanwhile, scammers are using AI to clone World Cup ticketing sites convincingly enough to fool fans, per WIRED. And Microsoft CEO Satya Nadella, writing for Seeking Alpha, argued the industry needs to stop fixating on building the largest models and start asking what those models actually accomplish for users.
What to Watch
The Kirkland-Palantir deal and 3M’s Ask 3M launch both point toward the same question: which industries will be first to see AI actually replace specialized human roles, rather than just assist them? Watch for announcements from other large law firms and manufacturers over the next few weeks — the partnerships being signed now will define who has an advantage in 2027. Nadella’s call to move beyond the frontier model race is also worth tracking; if Microsoft starts prioritizing application over raw capability, other major players may follow.