AI and the Future of Law: The 10 Year "Overnight" Success Story
HTML-код
- Опубликовано: 1 июн 2024
- Casetext started out in 2013 as a crowdsourced law library - a sort of “Wikipedia meets reddit” for the law. Ten years later, Casetext is one of the biggest wins to date in AI, capable of turning weeks of arduous legal work into hours or minutes. Just months ago it was acquired for $650 million dollars.
What happened between those two points?
For this episode of Main Function, YC President Garry Tan sits down with Casetext co-founder Jake Heller to learn the real story of their 10-year “overnight” success: the 3 a.m. origin story, how the company evolved as fast as tech would allow, and the “magic demo” that helped turn Casetext into a rocket ship.
Apply to Y Combinator: yc.link/MainFunction-apply
Work at a Startup: yc.link/MainFunction-jobs
Chapters (Powered by bit.ly/chapterme-yc) -
00:00 - Intro
00:48 - Part 1
03:09 - Part 2
10:56 - Part 3
17:14 - Outro 
Наука
What industry are you changing with AI?
Watching this is incredibly timely. We (Omnikey) applied for the W24 batch and are waiting to hear back. Such great info! As a lawyer, I’m really excited to see how AI positively influences the legal world and those who interact with it.
Service based businesses.
-Change- Not Create *New* !
I would have thought Y-Combinator would have got this already ?
Healthcare, in fact, building to brutally disrupt exiting traditional healthcare
Everyone who needs large GPT on prem, enclaved per department or workgroup, with quite lowcost inferencing clusters and rapid fine tuning. 1.2TB of unified VRAM for $75k in 2KW can do quite a bit for people who can't let their data leave the office.
"What a time to be alive."
ME: Wait, is this Y Combinator or Two Minute Papers?
I had thought saving lawyers time was a classic startup trap because law firms don't actually want to bill fewer hours.
😂
Suppose that if the delta is large enough, it forces their hand. "Use it or get left behind" kinda deal.
lololol
Not really, there is a curve for that. Better to have volume than the same grinding. PS: We do not our clients to go bankrupt. If we can save some money for them we do.
I think this is kind of an optimization problem, mixed with a "human factor": law firms definitely want to bill "enough or a reasonable number of hours" for a case, but also process any relevant evidence and do the legwork as fast as possible.
Chapters (Powered by ChapterMe) -
00:00 - Intro: The Story of Casetext
00:48 - Part 1: Jake's founder origins
03:09 - Part 2: How Casetext got to product market fit
04:36 - The Process
06:45 - Early access to GPT-4
08:52 - Golden demo
10:56 - Part 3: Moving fast and building the future
13:29 - Ex: California Innocence Project
14:25 - Mega opportunities for builders
16:08 - Never been a better time to start a company
17:14 - Outro
Iam working at a charterered accountant firm in india , i have to review the tax returns , most of the work is redundant and when i see technology like gpt it helps to reduce our work to a large extent , i literally feel like this technology will impact the world more than the internet doing provided if we can sort out hallucinations it has.
How many attempts, bro? I'm thinking of starting the CA exam at 25. (A lot of things have happened, and currently, I am trying to sort out my life.) Would you recommend the CA course at this age? Just a senior secondary (12th) pass-out here.
This is a great case study. Makes you think! Thank you for the push
Thank you Y combinator for sharing.
Thanks for sharing!
This is so inspiring
I applied to YC in September and I never got an email.
We are still reading and interviewing
Bought out by Thomson Reuters for 650M
What do you think are the downsides of a legal judgment prediction model? Do you think it is feasible in the real world?
i summarise the success of this company in 1 sentence
" Early access to gpt4 " . Like khan academy .
anything to do with altman's YC connections 🤔
still love your vidz tho
Doing RAG well is hard. Sanding down the edges is hard. Building a team is hard. Having proprietary data is hard. Selling is hard.
Good Job. adding International Legal requirements in our project. Custom requirements, Import/Export, product regulations, etc.
Is it me or was it a failing company until they connected to the OpenAI API
the past is in the past
No. They just reached a plateau of growth. The AI allowed them to accelerate their growth again.
Creating a viable business model around this idea seems challenging. The model would only be beneficial if it could incorporate oral supplements, but obtaining permission for audio file input, given that they are considered court property, is unlikely. Nevertheless, the model could still prove valuable for appeals or projects focused on innocence. The limitation lies in the financial aspect; if they had sufficient funds, they might have avoided unfavorable outcomes in the first place.
Disclaimer
This content was generated 40% by artificial intelligence. It may contain errors or inaccuracies, and should not be relied upon as a substitute for professional advice.
👏
Not an interview - it's an infomercial. Obvious omissions of any mention of hallucinations. It's all golden - I'll bet. CaseText and CoCouncil focus on establishment money. Not anything better.
Legal stuff in the US is strange. Here we just look up the law in the appropriate law book and the evidence.
If you're referring to the differences between common law and civil law, I find the fact that the court can technically be free to be entirely inconsistent and ignore precedent in civil law jurisdictions quite strange myself.