Harvey is valued at $11B. Legora just raised at $5.5B. I built their entire web application in two weeks and I'm making it open-source and free for everyone to use. Say hi to Mike: mikeoss.com. When I got the chance to try Harvey and Legora, I was surprised by how simple they were. A thought came to mind: I could probably build something similar in no time at all with Claude. And so I did. Assistant, project, tabular review and workflows. You get it all without vendor lock-in. Mike offers law firms an alternative, where they own the application layer and aren't stuck with a vendor they're renewing forever. You can try Mike in the demo on the website, or go to the GitHub link on the site to download the code and run a local version yourself.

Love this account

If you are a firm that hasn’t tried Claude or ChatGPT yet or started building your own harness, maybe consider it. Clearly Reddit has more negativity than what law firm surveys would reveal, but maybe these lawyers expressing their views under the safety of anonymity have a point when threads like this appear and trend on the Biglaw subreddit almost every month for the last two years. FWIW there are lawyers who I have met who genuinely like Harvey and Legora. But on the other hand typically their firms don’t provide access to alternatives.

There’s been quite a bit of speculation over whether Harvey and Legora would be acquired by the labs/hyperscalers. Many view it as the only plausible exit option for their investors given that if even OpenAI and Ant are having trouble with their IPO, it seems even more questionable that vertical AI companies with questionable margins would be able to. Let’s look at a few factors: 1. Revenue: in my view this is not a good justification for acquiring either. At around 500m and 200m “ARR” acquiring either would be an insignificant less than a percentage point increase to revenues for the Hyperscalers. I think general consensus is that these revenues are of questionable quality. And as I have posted before, I know of one V10 law firm which has run up a $2m Claude bill in a single month. So this one firm already generated >10% of Legora’s ARR for Ant. 2. Cash-flow: given that hyperscalers and labs are needing to constantly raise cash through debt and are seeking an IPO for further liquidity, acquiring Harvey and Legora would only exacerbate cash flow issues. Many of their current deals are on low fixed prices, or in some cases, completely free. Better to let them stay as independent entities that generate positive cash flow through token purchases funded by investors. 3. Distribution: both OpenAI and Ant have no trouble signing enterprise contracts with leading law firms and getting entreprise legal customers. In just the last two months, OpenAI has signed on a large portion of the leading v50 firms. I am hearing that more and more firms now have the full suite of Claude, ChatGPT, Harvey and Legora. 4. Data: This would assume that Harvey and Legora have captured valuable pre-training and RL data from their relationship with Biglaw in the last two years. Yet certainly in every contract, biglaw firms would have asked for a no model training clause. The cursor acquisition was the largest ever acquisition of a vertical application layer player by a hyperscaler. But that is justified by the massive user base and valuable training data it generates which SpaceXAI desperately needed to catch up to the frontier. 5. Model lock in: On the contrary, labs might view it as imperative to lock in the legal sector to their models. We have seen Harvey seek to transform itself into a neolab and going towards serving post-trained OSS models. This presents a potential risk to future revenues for the American closed model providers. They might thus take the view that acquisition might be the best way to stop this from playing out, and that they can worry about expanding revenues and margins later, so long as the lucrative Biglaw market is locked into OpenAI/Ant models through their control of the application layer.

As much as I agree that there is a massive ambition gap between America and the rest of the world, much of the phenomenon can be explained by the exorbitant privilege that the French president Valery Giscard d’Estaing pointed out. The US dollar being the reserve currency of the world allows the US to support its economy with monetary and fiscal measures beyond the reach of other economies. Countries with massive trade surpluses with America support the American economy by pumping their foreign reserves into the American capital markets. It’s easier to be ambitious when you have the strongest, most prosperous economy in the world, with the deepest capital and bond markets that can support risk taking impossible elsewhere. This then creates a virtuous cycle where talent, money and technology concentrates in America as it’s the place where capital gets the highest returns. The idea of the daring American and the indolent and risk averse European can therefore at least be partly explained by the economic structure of the world where American is the dominant economic hegemon. This inability to compete due to structural reasons then turned into cynicism over time. But remember that Europe was the economic hegemon with the greatest technology and scientists in the world until it lost its dominant position after two world wars.

Much of transactional work is based on a standard deal structure with templates and precedents. Attorneys are not figuring out a novel deal structure for a every deal. Many partners spend their whole lives twiddling around with the same set of documents that they have seen a thousand times and know almost every permutation of. Transactional work accounts for the majority of Biglaw revenues. Thus the unpredictability argument for the billable hour is not particularly strong.

Just realised that we reached 500 closed PRs on the Mike repo today: github.com/open-legal-produc… Maybe that's not a lot of for the token-maxxing demons running 20 sub agents and merging in 100 PRs per day, but it's been a lot of painstaking work from myself and the 20+ community contributors. It's been quite the journey learning to manage a viral open source project and I have learnt a lot along the way (waking up one day to 60+ PRs and 40+ issues after launching was certainly overwhelming). If you are interested in contributing, please do. We actively review PRs and issues and merge in contributions.

Two of the most elite firms on both sides of the pond, Wachtell and Slaughter and May, don’t use the billable hour.

Interesting slice of information here:

Many times you will see people question the veracity of something because it challenges their worldview. Instead of engaging in proper discussion first, the first thing they reach for out of instinct is a smug put down. It is neither polite nor does it reflect well on the person, as in doing so, they broadcast their ignorance to the world. I maintain the truth of this story, it came from a real conversation with a person I know.

This is brilliant

Yup

Proud to announce that lawve connector is now available in MikeOSS. Lawve.ai is one of the largest legal skills repositories in the word started by a French lawyer and ML PhD (@antoinelouis_ below). Via the connector you can now access their full set of legal skills written by lawyers and experts.

How do you steal a shophouse? The old guy has it in his name at the land registry. So it’s actually much more secure than crypto.

I know an associate in Biglaw who is a top 5 power user of AI at his firm. He’s not just making AI slop, he checks everything that the AI produces and verifies the all citations with actual sources, even though AI gets it correct most of the time these days. Recently he’s been told by his partner that, he’s “over confident”, that he’s working too fast and that he needs to check his work more, even though the partner could not pick out substantive issues in his drafts. I told him it’s likely code for: you’re running up a bunch of tokens and costing the firm while affecting my top line since I could normally bill the client twice for this work. As such, he is now forced to take longer on his work with little substantive impact on the work product while the client gets his bills multiplied. That’s probably part of the reason why despite the billions that have been invested in legal AI for Biglaw, there is barely any productivity increase and clients don’t see their bills go down by one cent.

ARR is so ingrained in tech that they have to annualise everything to make it sound bigger even if it doesn’t last for a year at all.

Sometimes it’s important to remember not to take the public opinion on this site too seriously. Remember a few months ago everyone was calling meta’s downfall and saying that Alexandr basically a dumbass and the whole scale AI buyout was a scam. Now it’s collective bootlicking, until meta falls behind on the benchmarks again and it’s back to public flogging.

Dont do it. The moment you do it you slouch forwards and start becoming spiritually old.