

Discover more from The Era of Generative AI
“Technology moves slowly, then all of a sudden.”
Hello all,
I’ve just finished Chairing the Generative AI Summit in London, where business leaders from across industries discussed the opportunities and risks these technologies present, and boy is GenAI shaking things up! I was struck by how quickly use cases are evolving: from designing NASA spaceships to writing product descriptions for lingerie, and powering new drug discovery. And speaking of industry disruption…
I hate reading contracts. So do most people. Fortunately, thanks to the staggering reach of generative AI, we may soon never need to read a contract again…
👨⚖️ Stephen Trusheim is the co-founder and CEO of Tome, a service that decodes legal contracts using AI. Stephen believes that everyone has the right to know what a contract means before they sign it. I don’t think that’s too big an ask, and it’s fascinating to see how GenAI is transforming the world of law.
In this episode we cover:
Stephen’s background [00:38]
The technology behind Tome [01:37]
Breakthrough moments in GenAI [03:24]
How ChatGPT raised AI’s “brand awareness” [05:41]
The pace of progress in AI [07:58]
Challenges of training large language models (LLMs) [08:42]
Red teaming [14:49]
Why contracts are dumb and unfair [19:14]
Automation vs Augmentation [21:46]
The future of contracts with Tome [27:41]
A somewhat pessimistic vision of the future… [30:05]
Whether corporate monopolies will dominate [32:08]
Training data, fair use and current lawsuits [38:10]
Choosing our future and prioritizing liberty [41:44]
Solving a social need
How can tools like ChatGPT solve real, human problems? Stephen shared a shocking(?) statistic: 94% of Americans cannot afford a lawyer.
“The gap between the services people can actually afford, and the problems they face in real life, is huge.”
Considering this reality, Stephen sees the power of AI in expanding the set of people who can afford these vital services.
What does this mean for lawyers?
Will my job be automated? Is this about automation or augmentation? Who benefits?
For Stephen at Tome, he does not mince words: “We automate lawyers. That’s what we do.”
The reality is that AI is good at automating-away rote, white-collar tasks. These technologies are disrupting jobs right across the skill spectrum. Adapt, adapt, adapt…
“I think AI is going to have tremendously negative effects on society if we don’t do it right.”
Let’s get this one right. There’s a great deal hanging in the balance if we don’t.
Namaste,
Nina
PIONEER: Decoding Contracts with AI | Stephen Trusheim
Improving access to legal expertise is one of the best applications of GenAI I've heard of, especially considering the cost of a lawyer in the USA. However, a legal professional I spoke with at the Generative AI Summit last week raised an interesting point: the first several years of a (human) lawyer's career is spent doing rote tasks, before they are considered experienced enough to tackle more specialist cases. By automating the "basics", we remove this traditional learning path, potentially making it more difficult to become a lawyer, which could make accessing legal services even harder. If we could automate all legal proceedings at all levels then this problem is irrelevant, but until this is possible perhaps the legal sector should be preparing to adapt the route into their profession.
The less lawyers in the world the better