It seems that a lot of people would rather accept a relatively high risk of unfair judgement from a human than accept any nonzero risk of unfair judgement from a computer, even if the risk is smaller with the computer.
GPT-5 outperforms federal judges in legal reasoning experiment
181–190 of 254 posts
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#182Generative AI is not making judgements or reasoning here, it is reproducing the most likely conclusions from its training data. I guess that might be useful for something but it is not judgement or reasoning.
What consideration was given to the original experiment and others like it being in the training set data?
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#183Earlier quoted context omitted.
Whole "democracy" thing is legal framework that wealthy and powerful people built to make safe wealth transfer down the generations possible while giving away as little as possible to average joe. In a countries without this legal framework its usually free for all fight every time ruling power changes. Not good for preserving capital. So wealthy having more rights is system working as intended. Not inherently bad th…
>"So wealthy having more rights is system working as intended. Not inherently bad thing either" Sorry but I do not feel this way. "Not inherently bad thing either" - I think it is maddening and has to be fixed no matter what. You know, wealthy generally do not really do bad in dictatorial regimes either.
Until they found dead with unexpected heart attack, their car blow up or they fall out of the window.
In dictatorship vast majority of wealthy people no more than managers of dictators property. Usually with literal golden cages that impossible to sell and transfer.
Once person fall out of favor or stop being useful all their "wealth" just going to be redistributed because it was never theirs.
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#184Earlier quoted context omitted.
Yeah, I'm reminded of the various child porn cases where the "perpetrator" is a stupid teenager who took nude pics of themselves and sent them to their boy/girlfriend. Many of those cases have been struck down by judges because the letter of the law creates a non-sequitur where the teenager is somehow a felon child predator who solely preyed on themselves, and sending them to jail and forcing them to sign up for a se…
This example feels more like a bug in the law itself that should be corrected. If this behavior is acceptable then it should be legal so we can avoid everyone the hassle in the first place. I bet AI would be great at finding and fixing these bugs.
One might imagine a distant future where laws could be dramatically simplified into plain-spoken declarations, to be interpreted by a very advanced (and ideally true open source) future LLM. So instead of 18 U.S.C. §§ 2251–2260 the law could be as straightforward as:
"In order to protect children from sexual exploitation and eliminate all incentive for it, no child may be used, depicted, or represented for sexual arousal or gratification. Responsibility extends to those who create, assist, enable, profit from, or access such material for sexual purposes. Sanctions must be proportionate to culpability and sufficient to deter comparable conduct."
...and the AI will fill in the gaps.
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#185Earlier quoted context omitted.
Yeah, I'm reminded of the various child porn cases where the "perpetrator" is a stupid teenager who took nude pics of themselves and sent them to their boy/girlfriend. Many of those cases have been struck down by judges because the letter of the law creates a non-sequitur where the teenager is somehow a felon child predator who solely preyed on themselves, and sending them to jail and forcing them to sign up for a se…
This is one of the roles of justice, but it is also one of the reasons why wealthy people are convicted less often. While it often delivered as a narrative of wealth corrupting the system, the reality is that usually what they are buying is the justice that we all should have. So yes, a judge can let a stupid teenager off on charges of child porn selfies. but without the resources, they are more likely be told by a p…
A teenager posting own photo and getting away with it is massively different then a rich guy raping a girl and getting away with it. Or, rich guy getting away with outright frauds with thousands of victims.
> While it often delivered as a narrative of wealth corrupting the system, the reality is that usually what they are buying is the justice that we all should have.
This is not true. Epstein did not got "justice we all should have". Trump did not got "justice we all should have". People pardoned by Trump did not got "justice we all should have". Wall Street and billionaires are not getting justice we all should have either. All these people are getting impunity and that is not what we all should have.
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#186Earlier quoted context omitted.
> It was regarded as a model for other states. Really? That "model" has the common, but obviously extremely undesirable, feature of criminalizing sexual relationships between students in the same grade that were legal when they formed . How could it be regarded as a model for anyone else?
You might have misread it. Texas' model is decriminalizing teens sexting, not criminalizing it
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#187IANAL, but this seems like an odd test to me. Judges do what their name implies - make judgment calls. I find it re-assuring that judges get different answers under different scenarios, because it means they are listening and making judgment calls. If LLMs give only one answer, no matter what nuances are at play, that sounds like they are failing to judge and instead are diminishing the thought process down to black-…
Unfortunately, as the aptly titled 'Noise' [1] demonstrated o so clearly, judges tend to make different judgement calls in the same scenarios at different times.
1. Noise - https://en.wikipedia.org/wiki/Noise:_A_Flaw_in_Human_Judgmen...
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#188The problem is that biases tend to be built in via even rudimentary stuff like bad training material and biased tuning via system prompts. E.g., consider the 2026 X post experiment, where a user ran identical divorce scenarios through ChatGPT but swapped genders. When a man described his wife's infidelity and abuse, the AI advised restraint to avoid appearing "controlling/abusive." For a woman in the same situation,…
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#189Earlier quoted context omitted.
You might have misread it. Texas' model is decriminalizing teens sexting, not criminalizing it
I didn't misread it, but apparently you did. Why is criminalizing an existing legal relationship a good idea?
Re: GPT-5 outperforms federal judges in legal reasoning experiment
#190Earlier quoted context omitted.
This is one of the roles of justice, but it is also one of the reasons why wealthy people are convicted less often. While it often delivered as a narrative of wealth corrupting the system, the reality is that usually what they are buying is the justice that we all should have. So yes, a judge can let a stupid teenager off on charges of child porn selfies. but without the resources, they are more likely be told by a p…
> it is also one of the reasons why wealthy people are convicted less often. A teenager posting own photo and getting away with it is massively different then a rich guy raping a girl and getting away with it. Or, rich guy getting away with outright frauds with thousands of victims. > While it often delivered as a narrative of wealth corrupting the system, the reality is that usually what they are buying is the justi…
The pardons (the non-purchased ones) were not out of charity to the pardonees but to foster future behavior beneficial to the pardoner.