Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A.I. note takers are making lawyers nervous

nytimes.com

Post reply on HN
Print view View original

191–200 of 204 posts

Re: A.I. note takers are making lawyers nervous

#191

Post by lesostep »

feoren wrote: post #76
r_lee wrote: post #29

Earlier quoted context omitted.

I don't think it's a training issue, it's simply that there's no inherent "I don't know" in the transformer architecture unless it's really like something completely unknown, otherwise the nearest neighbor will be chosen and that will be whatever sounds similar or is relevant, even if it might cause a problem

The final output of the neural network part of an LLM is a vector with weights for every token, that is then usually softmaxed and picked from. Can we not quantify the uncertainty by looking at the distribution of weights of the top 10 options? Like we expect for a note-taking app that the top choice would be something like 98% certain, and if we see that the model gives a weight of 60% to "Russia" and 30% to "France…

I think it might break the game. Most words sound similar enough to other words. "cat" and "get", "he simply" and "his simply", etc.

Add accents, and half the words would be indistinguishable from each other (note that word "indistinguishable", ironically, would be quite distinguishable).

People parse things like that in so much context, based in their own understanding of a situation, their grasp on speakers accent or speech impairments, etc.

Add to that that most native english speakers blur words together. The pause that in some languages is used to separate words, is used in english to separate sentences. English language as spoken doesn't separate words natively.

The text-to-speech before LLMs was meh. I think it's the ability to generate filler for uncertain words that makes it feel magic compared to before.

Re: A.I. note takers are making lawyers nervous

#192

Post by flumes_whims_ »

moffkalast wrote: post #35
user_7832 wrote: post #14

Earlier quoted context omitted.

> If they miss a word they never do unintelligible, they just start playing madlibs based on the rest of the sentence. Imo this is the single biggest flaw of LLMs. They're great at a lot of things, but knowing when they're wrong (or don't have enough information to actually work on) is a critical flaw. IMO there's nothing structural about why they shouldn't be able to spot this and correct themselves - I suspect it's…

It's a benchmark and eval issue. Guessing gets them the right result sometimes and the models rank better in error rate than they'd otherwise. We need the kind of benchmarks that penalize being wrong WAY more than saying "I don't know". Of course there's a secondary problem that the model may then overuse the unintelligible option, but that's something that's a matter of training them properly against that eval. You…

Benchmarking for giving I don't know rather than wrong answer seems to be the right way to steer industry towards making models that are good at this. AA-Omniscience is one such benchmark.

AA-Omniscience is a knowledge and hallucination benchmark that rewards accuracy, punishes bad guesses and provides a comprehensive view of which models produce factually reliable outputs across different domains. The benchmark contains 6,000 questions across 6 major domains, derived from authoritative academic and industry sources and generated automatically using an LLM-based question generation agent to ensure unambiguity, scalability and factual precision

https://artificialanalysis.ai/evaluations/omniscience

Re: A.I. note takers are making lawyers nervous

#193

Post by flumes_whims_ »

nsvd2 wrote: post #107
xoxxala wrote: post #100

Earlier quoted context omitted.

When I say something inappropriate on the phone or over Discord, I always tell the NSA guy listening it was a joke.

This is an old joke, but as AI becomes more and more capable it becomes increasingly likely that the government will have some kind of intelligence sifting though every post you make and message you send.

And we can guess that recordings for have stored for years back before it got feasible to actually process.

Re: A.I. note takers are making lawyers nervous

#194

Post by sdellis »

cj wrote: post #170
sdellis wrote: post #164

Earlier quoted context omitted.

You are making a lot of assumptions about the "average" person. Here's a Pew Research study that says otherwise: https://www.pewresearch.org/internet/2019/11/15/americans-an...

That research study also concludes that “59% of people have no understanding of what companies do with the data they collect” To me, that says the average person doesn’t care enough (they care, just not enough) to do anything about it. They might care enough to spend 5 seconds signing a petition, but not enough to spend 5 minutes installing an ad blocker, and definitely not enough to spend 5 hours doing anything more…

Look, no offense, but I'm gonna trust the peer reviewed research article over your hot take.

Re: A.I. note takers are making lawyers nervous

#195

Post by user_7832 »

cwillu wrote: post #125
user_7832 wrote: post #105

Earlier quoted context omitted.

Damn, did I say something wrong or unpopular to get a downvote?

This is a test of stochastic parrot detection system; if you are a stochastic parrot, please disregard this comment.

I am too sleepy to understand this comment :(

Do stochastic parrots dream of the number of 'e's in "electric sheep"?

Re: A.I. note takers are making lawyers nervous

#196

Post by cj »

sdellis wrote: post #194
cj wrote: post #170

Earlier quoted context omitted.

That research study also concludes that “59% of people have no understanding of what companies do with the data they collect” To me, that says the average person doesn’t care enough (they care, just not enough) to do anything about it. They might care enough to spend 5 seconds signing a petition, but not enough to spend 5 minutes installing an ad blocker, and definitely not enough to spend 5 hours doing anything more…

Look, no offense, but I'm gonna trust the peer reviewed research article over your hot take.

No offense taken. This is just a friendly debate.

Re: A.I. note takers are making lawyers nervous

#197

Post by keernan »

LanceH wrote: post #32
vintagedave wrote: post #2

Paywall: can anyone share what the issue is? Inaccuracy in meeting minutes? Leaking private info, re security of notes? I have never used them (don't trust them to accurately capture what is important in a meeting vs just noting what's mentioned), but the concept seems very useful to me.

They are saying that it could invalidate attorney client privilege because the transcription could technically be available to an outside party. I suspect what isn't being said by the lawyers is they want to keep attorney client privilege so they can outright lie.

>they want to keep attorney client privilege so they can outright lie

As a trial attorney for over 40 years, that is an incredibly offensive take. Attorney/client privilege is usually litigation related and a prime example of the nature of conversations involve our advising our client of the prospects of prevailing at trial and whether to engage in settlement discussions with money amounts involved. If some day you are sued and you have a conversation with your lawyer about your financial worth as well as how much you are willing to pay to the person suing you - and that information ends up being turned over to the person suing you - you won't be so snide about the importance of attorney client privilege.

Re: A.I. note takers are making lawyers nervous

#199

Post by darthwalsh »

sandworm101 wrote: post #13

>> Executives and corporate boards generally expect conversations with their legal team about legal matters to have attorney-client privilege. They lose that protection if they share the same information with outside parties — and it’s possible that an A.I. note taker could have the same effect. Total oversimplification. The fact is the privilege is a rule totally in the hands of the court. Every time a new communica…

Thanks for sharing that nuance. It seems one court weighed in that Claude chat wasn't accepted:

> In February 2026, Judge Jed S. Rakoff of the U.S. District Court for the Southern District of New York ruled that a defendant who pasted information—including details conveyed by his lawyers—into a public, consumer-grade AI chatbot (specifically Anthropic's Claude) completely waived his attorney-client privilege.

But maybe if you are using a transcription tool that happens to send your audio through the Anthropic APIs, the ruling would be different?

Re: A.I. note takers are making lawyers nervous

#200

Post by Semaphor »

netdevphoenix wrote: post #190
Semaphor wrote: post #95

Earlier quoted context omitted.

I often (ish) translate between English and German, two languages I speak very well. The quality of translation is amazing and far better than what old systems did. Maybe it depends on topics or length, for me it's usually 1-2 paragraphs of a German article to share online.

> The quality of translation is amazing and far better than what old systems did. Are you native in both languages? If you are only native in one of them, it would be insightful to find if people with your skillset but native in the language you are not have the same opinion as you.

It’s rather unlikely that the translation in one direction is great, but lacking in the other, while also being just good enough (compared to before) that my close-to-native English skill misses it, while the old google translate somehow magically made me think it was bad.

Sadly there are no examples here to compare.

Post reply on HN