Rise of the Machines

[Politesse]

But information websites are exactly what AI trawls to get its information...

 

[Elixir]

My expectation is not that we will see a uniform stupification of the populace,

Stupefaction, surely?

Perfect. Bang on their heads until their brains are semi-liquid mush.

 

[Elixir]

I see this as a good thing, to see all the random ass clowns spewing unvetted information for the sake of clicks and ad revenue being burned out of existence.

I’ve had that thought, but you said it better!

 

[bilby]

It Was Only a Matter of Time: *Both* Sides’ Lawyers Sanctioned for Using AI

According to the AI Hallucination Database, there have been 1,598 cases to date, worldwide, in which someone got in trouble for relying on a generative AI tool that turned out to have just made stuff up.

Of course I haven’t read all of those, but I suspect this might be the first time that all the lawyers on both sides were kicked off the case for doing that.

...

“This Court is yet again burdened with addressing AI hallucinations in court filings,” the opinion says, citing a number of prior cases. “It has previously acknowledged that AI is a powerful tool that when used prudently, provides immense benefits.” (Emphasis in original, scoffing at the word ‘immense’ added by me.) “However,” the court continues, while generative technology can produce words, it cannot attach sincerity, truth, or responsibility to what it writes.”

 

[Colonel Sanders]

It Was Only a Matter of Time: *Both* Sides’ Lawyers Sanctioned for Using AI

According to the AI Hallucination Database, there have been 1,598 cases to date, worldwide, in which someone got in trouble for relying on a generative AI tool that turned out to have just made stuff up.

Of course I haven’t read all of those, but I suspect this might be the first time that all the lawyers on both sides were kicked off the case for doing that.

...

“This Court is yet again burdened with addressing AI hallucinations in court filings,” the opinion says, citing a number of prior cases. “It has previously acknowledged that AI is a powerful tool that when used prudently, provides immense benefits.” (Emphasis in original, scoffing at the word ‘immense’ added by me.) “However,” the court continues, while generative technology can produce words, it cannot attach sincerity, truth, or responsibility to what it writes.”

I personally know an attorney who, along with two of her colleagues, were sanctioned for the use of A.I. in a brief. Their firm had subbed the job out to an outside lawyer and that brief was used in court. No one reviewed it for accuracy and its appropriateness with respect to the parties and facts of the case.

I once tried a free trial from an A.I. firm. I'd already written the brief for this particular case and submitted it, but I wanted to see how this A.I. program would compare.

It was absolute shit. The only way I can see this ever working in my field is for the program to be able to interview clients, read and interpret earlier court rulings, oppositions arguments, and have an attorney input all the facts of the case and why our argument is better.

Put another way, doing all the work to hope that the AI gets it right would take just as long, if not longer than actually doing the research and writing yourself.

It may be good for finding nuggets of info in thousands of pages of discovery, but the thing about that is that the AI's data, along with all those pages of discovery is still going to have to be reread and interpreted for accuracy.

"I didn't fuck it up, the AI did!" will never, ever, ever, ever, never be a valid excuse in a courtroom.

 

[thebeave]

It Was Only a Matter of Time: *Both* Sides’ Lawyers Sanctioned for Using AI

According to the AI Hallucination Database, there have been 1,598 cases to date, worldwide, in which someone got in trouble for relying on a generative AI tool that turned out to have just made stuff up.

Of course I haven’t read all of those, but I suspect this might be the first time that all the lawyers on both sides were kicked off the case for doing that.

...

“This Court is yet again burdened with addressing AI hallucinations in court filings,” the opinion says, citing a number of prior cases. “It has previously acknowledged that AI is a powerful tool that when used prudently, provides immense benefits.” (Emphasis in original, scoffing at the word ‘immense’ added by me.) “However,” the court continues, while generative technology can produce words, it cannot attach sincerity, truth, or responsibility to what it writes.”

I personally know an attorney who, along with two of her colleagues, were sanctioned for the use of A.I. in a brief. Their firm had subbed the job out to an outside lawyer and that brief was used in court. No one reviewed it for accuracy and its appropriateness with respect to the parties and facts of the case.

I once tried a free trial from an A.I. firm. I'd already written the brief for this particular case and submitted it, but I wanted to see how this A.I. program would compare.

It was absolute shit. The only way I can see this ever working in my field is for the program to be able to interview clients, read and interpret earlier court rulings, oppositions arguments, and have an attorney input all the facts of the case and why our argument is better.

Put another way, doing all the work to hope that the AI gets it right would take just as long, if not longer than actually doing the research and writing yourself.

It may be good for finding nuggets of info in thousands of pages of discovery, but the thing about that is that the AI's data, along with all those pages of discovery is still going to have to be reread and interpreted for accuracy.

"I didn't fuck it up, the AI did!" will never, ever, ever, ever, never be a valid excuse in a courtroom.

I assume you probably heard about this. AI has its place when used appropriately and cautiously, but it certainly does not seem ready for prime when it comes to legal matters:

California Supreme Court Seeks Answers About Bar Exam’s Use of AI

 

[Bronzeage]

For better or worse a phenomenon called zero-click is causing informational website traffic to nosedive causing marketers to rethink where they spend their ad dollars. If everyone can get their answer from AI, why hunt through a website searching for that same information? Bloomberg and Forbes articles state that roughly 60% of searches now result in no clicks at all. Tentatively, I see this as a good thing, to see all the random ass clowns spewing unvetted information for the sake of clicks and ad revenue being burned out of existence. The survivors will be government research, charities, encyclopedias, archives, e-learning, etc.

Many years ago, Doonesbury predicted this. In the scenerio, Doonesbury was a Tech Bro, before there was such a thing. There was a "visionary" on the staff, whose job was to think of new things. His big idea was a computer program for a shopper that would survey the entire Internet and find the best price for an item. Doonesbury considered it and told him ads are revenue for web sites and his idea meant no one looking at ads, and no eyes on ads meant bankruptcy. Long before anyone ever heard of AI.

 

[pood]

I really like AI. Not.

Two of my favorite American writers are Cormac McCarthy and Gore Vidal. They’re both dead. I’ve no idea what they thought of each other, or whether they even read each other.

So I decided to ask Google AI. What a stupid thing to do!

Well, it turns out that McCarthy disliked Vidal’s work, because he believed it was too “solipsistic and mannered.” Huh. This immediately got my antennae up, because I know that the famously terse and taciturn McCarthy did NOT speak or write like that. Helpfully, shitty Google AI did provide a link for the source of the “quote.” What was it?

It turned out to be Jerry Coyne’s Why Evolution Is True blog, which I visit every day, mainly to scream at Coyne’s political views, especially his support of genocide in Gaza and his rooting on of Trump’s war in Iran. I do like his biology stuff.

Now why did Google AI link to Coyne’s blog?

I guess because it is tracking my online movements.

The linked post in question was about some list or other of the ten greatest American novels of the 20th/21st century. Both McCarthy and Vidal were on the list.

But it was Jerry Coyne who dismissed Vidal as too “solipsistic and mannered,” and NOT Cormac McCarthy!

Thanks ever so much for your “help,” Google Artificial “Intelligence.”

Next I trust Google AI will helpfully tell me what Vidal thought about McCarthy!

Jesus, we are careening toward the abyss.

 

[Michael S. Pearl]

I really like AI. Not. ... Thanks ever so much for your “help,” Google Artificial “Intelligence.”

Once you realize that the 'A' in "AI" stands for "Alleged", then you will understand all that you need to understand.

 

[pood]

So, I asked Google Alleged Intelligence what Vidal thought of McCarthy. Here is what it gave me:

Gore Vidal was largely unimpressed by Cormac McCarthy. He found McCarthy's early work "interesting" but grew critical of his later writing. Vidal was dismissive of the Border Trilogy, calling it "progressively worse and more and more sentimental," and actively avoided reading The Road after sampling passages from it.

Huh.

But it gave me a link! Back to Jerry Coyne!

And the above quote is neither from Vidal nor Coyne, but from a poster on Coyne’s blog named Tim Harris! It is Tim Harris’s opinion!

Just shoot me now.

 

[Colonel Sanders]

It Was Only a Matter of Time: *Both* Sides’ Lawyers Sanctioned for Using AI

According to the AI Hallucination Database, there have been 1,598 cases to date, worldwide, in which someone got in trouble for relying on a generative AI tool that turned out to have just made stuff up.

Of course I haven’t read all of those, but I suspect this might be the first time that all the lawyers on both sides were kicked off the case for doing that.

...

“This Court is yet again burdened with addressing AI hallucinations in court filings,” the opinion says, citing a number of prior cases. “It has previously acknowledged that AI is a powerful tool that when used prudently, provides immense benefits.” (Emphasis in original, scoffing at the word ‘immense’ added by me.) “However,” the court continues, while generative technology can produce words, it cannot attach sincerity, truth, or responsibility to what it writes.”

I personally know an attorney who, along with two of her colleagues, were sanctioned for the use of A.I. in a brief. Their firm had subbed the job out to an outside lawyer and that brief was used in court. No one reviewed it for accuracy and its appropriateness with respect to the parties and facts of the case.

I once tried a free trial from an A.I. firm. I'd already written the brief for this particular case and submitted it, but I wanted to see how this A.I. program would compare.

It was absolute shit. The only way I can see this ever working in my field is for the program to be able to interview clients, read and interpret earlier court rulings, oppositions arguments, and have an attorney input all the facts of the case and why our argument is better.

Put another way, doing all the work to hope that the AI gets it right would take just as long, if not longer than actually doing the research and writing yourself.

It may be good for finding nuggets of info in thousands of pages of discovery, but the thing about that is that the AI's data, along with all those pages of discovery is still going to have to be reread and interpreted for accuracy.

"I didn't fuck it up, the AI did!" will never, ever, ever, ever, never be a valid excuse in a courtroom.

I assume you probably heard about this. AI has its place when used appropriately and cautiously, but it certainly does not seem ready for prime when it comes to legal matters:

California Supreme Court Seeks Answers About Bar Exam’s Use of AI

I didn't know about that. I don't practice anymore and passing the bar was more like surviving an airplane accident than accomplishing something worthwhile.

Anyway, I don't see how AI could screw this up more than the humans who design the multiple choice section of the exam. The bar exam should be difficult, but good god. Most of the questions are so vague and convoluted that the right answer is often the least wrong or the best of four marginally right answers.

 

[Elixir]

I don't see how AI could screw this up more than the humans who design the multiple choice section of the exam. The bar exam should be difficult, but good god. Most of the questions are so vague and convoluted that the right answer is often the least wrong or the best of four marginally right answers.

Well, it’s a very similar deal with AI. The language model offers billions of possible responses, and narrows it down by frequency of similar responses to similar queries. It ends up selecting complete answers that come from a mix of actual facts, parody, facetious comments etc, often rendering responses weirdly incoherent upon close examination.

 

[thebeave]

My next question to AI is to ask it how many fingers there are on a human hand. If it responds with six, then I'm checking out of it for good.

 

[Elixir]

My next question to AI is to ask it how many fingers there are on a human hand. If it responds with six, then I'm checking out of it for good.

Gimme six!

 

[pood]

I just want to know if AI can tell me how many letters “r” there are in “strawberry.”

 

[Elixir]

I just want to know if AI can tell me how many letters “r” there are in “strawberry.”

 

[Loren Pechtel]

Put another way, doing all the work to hope that the AI gets it right would take just as long, if not longer than actually doing the research and writing yourself.

Which is how it is in every skilled field.

It may be good for finding nuggets of info in thousands of pages of discovery, but the thing about that is that the AI's data, along with all those pages of discovery is still going to have to be reread and interpreted for accuracy.

Exactly. It's real value is as a fuzzy search engine. Analytic works, generative does not.

 

[Politesse]

We're being pushed so hard right now to teach the students "how to use AI responsibly". But if our own professional opinion is that AI simply isn't a responsible tool for this task or that, we get a sit-down with the dean about embracing change and preparing students for the future.. It all feels very gaslit.

 

[Tigers!]

So, I asked Google Alleged Intelligence what Vidal thought of McCarthy. Here is what it gave me:

Gore Vidal was largely unimpressed by Cormac McCarthy. He found McCarthy's early work "interesting" but grew critical of his later writing. Vidal was dismissive of the Border Trilogy, calling it "progressively worse and more and more sentimental," and actively avoided reading The Road after sampling passages from it.

Huh.

But it gave me a link! Back to Jerry Coyne!

And the above quote is neither from Vidal nor Coyne, but from a poster on Coyne’s blog named Tim Harris! It is Tim Harris’s opinion!

Just shoot me now.

You are in the land paved with guns and you cannot find anyone to shoot you?
/sarcasm