
WillC
373 posts

WillC
@willchen500
Oxford law, former Latham associate


The Thomson Reuters 2026 AI in Professional Services report is not super bullish for the company or AI's progress in legal overall. badfaith.beehiiv.com/p/ai-roi-more-…










I think flaming is generally bad taste, but it’s also true AI affords us a level of abstraction where a real human job is developing normative takes about the technologies we pursue. in this screenshot thread I noted there’s persistent sense across the timeline that Harvey is a meme play for Boomers who want Secure Chat. @scottastevenson appropriately calls this Champagne AI. My smartest friends seem to want to work on AI plays that exploit some property of AI to introduce a new value proposition, structurally counter an old one, or expand service to an underserved audience. Harvey does none of these. It’s at best an expensive lifestyle product for the well heeled partner class. That seems to make for a good business, but we shouldn’t pretend that’s particularly interesting. AI has a bad reputation in part because it’s a product specifically for Capital. Harvey is a canonical example of this. Does Harvey make legal more accessible? Or change the business model of legal? No. No Sequoia think piece will change this. A partner at a Legora customer excitedly shared with me how they were using AI at the firm. I asked > cool so what’s in it for me > cheaper service? 😉 I got no answer. As AI races up the stack, discussion of norms and aesthetics is exactly the sort of human exercise we should expect to do more. Raising Prestigious Capital doesn’t excuse you from being responsible to questions like these — if anything it raises the expectations to have clear answers.



How does a seasoned Supreme Court lawyer prepare for the biggest case of his life? Using Harvey. Read how Harvey supported @neal_katyal in refining his arguments before the Supreme Court and how we are bringing those tools to law schools with Harvey Moot: harvey.ai/blog/the-supre…

Smart use of AI, so why lie about it? Why call Harvey a “bespoke AI I built over the last year with a legal AI company”? If he used CoCounsel or Claude for the same tasks, would he say he built Westlaw or Anthropic?




I’m not an investor in Harvey or Legora but I find it weird that all the hate has been directed towards Harvey when Legora is MUCH more hate-worthy Alright, flight taking off now, see y’all in a bit

Introducing 5000+ legal agents on Mike


I was interviewed yesterday by The Australian Financial Review on Mike. The article notes the game-changing impact that Mike has had on the legaltech industry, less than a week after release. The article also features the first public comment from Harvey. Their spokesperson stated that "big proprietary platforms such as Harvey remained best placed to meet the intensive technology needs of law firms, including robust data security, around-the-clock support and access to a range of large language models." I agree on the security and service requirements of big law firms. But let me address the "proprietary" bit. A piece of software being private rather than open source does not equate to having technological moat, nor does it mean that it is secure. The idea that private = secure and open source = insecure is a pretty widespread misconception in the legal industry. Some of most secure and robust software in the world used by everyone, like Linux, is open source. Any modern piece of software is built upon the foundation laid by open source libraries, and that includes Harvey. It is precisely because of open source communities that we have these public goods. Article: This ‘game changer’ free app could blow up the $23b legal AI sector afr.com/companies/prof…



