Related Posts
**Thoughtworks Referrals**
Quality Analyst (4+ years exp in API automation)
Location : Bangalore,Gurgaon , Hyderabad, Mumbai , Coimbatore
Senior Java developer(4-9 yrs exp)
Location: Pune, Mumbai ,Gurgaon , Bangalore , Hyderabad, Chennai
Senior data engineer(4-9 yrs exp)
Location: Pune, Mumbai ,Gurgaon , Bangalore , Hyderabad
If you are seriously looking for switch, Mail your resume to devm8366@gmail.com for referral, with preferred location
Anyone from London ?
Additional Posts in Law
Pro
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
Rising Star
Have you read the latest research on AI and productivity? Lookup “ai productivity paradox.” It’s indisputable that AI has the *potential* to be transformative but we’re actually not that close to seeing massive economic impact from AI, especially not in law. Basically, we’re seeing productivity gains at the individual/task level for knowledge workers, but not so much at the company/aggregate level (and some companies actually see productivity *losses* before seeing any gains). And that’s doubly true in legal, where ai adoption has been measured and slow. Maybe I’m missing something but I don’t think we’re 5 years from significant cuts, at least not with the current state of the tech. What’s actually more likely in the near term is that workloads initially INCREASE as AI adoption accelerates and generates MORE demand in professional services (at least initially). also, keep in mind there will still a lot of slop to be contended with while the tech matures, and someone has to catch/correct it.
And a study from last year showing that programmers given access to AI tools THOUGHT they were working faster, but were actually working slower. Now, I don’t know if that was large enough to be statistically accurate, and it can’t possibly have been a double blind study (except in the sense that each group might not have known they were being compared to the performance of another group), and programming speed and quality absolutely depends on each individual’s ability and knowledge…but it squares with what I’ve seen elsewhere. As an attorney, I can tell you it takes me MORE time to review and edit a document prepared by someone else than it would to just prepare it myself.
No one should trust AI to do real legal work.
I don’t worry about AI at all.
I have a niche in my firm where I do almost exclusively dispositive and pre-trial motions. AI is not going to be able to do that. The point of that is coming up with arguments no one has made before. AI doesn’t really do that. I don’t think it’s controversial to say AI is good at creating images and text, but it doesn’t come up with good new ideas.
Another big problem is that, at this point, AI is accurate maybe 90% of the time. In my book, that means you can trust it 0% of the time.
I’m sure AI has some good uses, like letters or perfunctory motions. But people should have templates for those types of things. I don’t think it would take that much longer than AI.
I always tell people AI will never take my job because people are infinitely stupid and it requires infinite creativity to get them out of their messes.
There's a huge divide in partners' views of what AI can do. Had a partner say to me the other day that they foresee AI eliminating the need for juniors and younger attorneys, and partners or senior attorneys will be able to do it all for the most part using AI. It showed me how little this person actually understands about the capabilities of AI (and, perhaps more importantly, their ego in thinking they could service clients without support from others). Then you have other partners who've never used Harvey for even the most basic of tasks because they're set in their ways and not looking to change. Either way, the fact that you have such wildly divergent perspectives existing in the same firm means we're a lot more than 5 years out from a sea change and drastic reduction in the number of attorneys. Those who don't get proficient using AI will absolutely be left behind by those who incorporate it as a tool in their arsenal, but it's not the complete doomsday for the legal profession folks like to talk about on this app.
A4 is that happening or is that the excuse? Fair? No, nothing is fair. But it’s a headcount business so that’s why I don’t see it changing soon.
AI is just giving me more work to do. Everyone with an axe to grind is using it to represent themselves. The papers are horrible. My firm does education law, so we run into a lot of pro se parties. People who previously would not represent themselves are all now using AI, and the crackpots are even crazier.
I would strongly advise a shift in the mindsets I'm seeing in this chain...
I now pay for 2 OC hours to review my team's own AI-supported internal analysis to get a *better* result in tasks that would've been 10+hrs of OC billables before. The outputs much more grounded in our company-specific realities because that's where my team - not OC - is the SME. Even for non-corporate tasks, realize that AI quality is evolving faster than you can imagine: in my experience, AI response sophistication has improved 30-40% over the past 12 months alone. If you're not using it regularly, you wouldn't even know. You do the math on the reduced billable demand potential - but this will hit VERY unevenly at first depending on specialization, client type, industry, etc.
I get that YOU don't trust AI, but in short order that obstinacy and refusal to adapt isn't going to land well with the people paying your invoices (me).
I guess I don’t disagree with most of this. I’m curious what kind of work product you’re talking about.
I think your threat at the end should be a little more narrowly tailored. My firm gets hired because we win more trials than anyone else. We’re not exactly gonna be dropping our billing by 80% in response to AI. See what happens if you go with the firm that does. And also the firm that lets themselves be micromanaged.
Not saying you’re necessarily doing it here, but I’m not on board with the fetishization of AI. It’s something very powerful.
But it’s not *intelligence* in the way we would normally describe it. Human input will be required for a long time.
I didn’t think big 4 had NEP?
Oh, I meant associate 3* (are you okay?)
F
I would say we have at least 10 years before AI seriously disrupts legal hiring. Even the best legal-tailored AI tools hallucinate 15-20% of the time. Moreover, accurate OCR for using AI in transactional work with .pdf's is a problem that isn't closed to being fixed. It will get better, but the marginal improvements from hereon forward are more difficult than the big gains made in the last few years. It will impact new law grads for sure, but midlevel and senior attorneys will still be necessary, perhaps even moreso, to correct AI's work and exercise human judgment. Take a look at all the cases on sanctions for attorneys and pro se litigants relying on AI to get a good idea of AI's shortcomings.
Is it job.