Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror

Comment Catch - 0001 0110 (Score 1) 41

One example shared in the patch shows a simple fix to a typo in the kernel's OPP documentation. Claude, an AI assistant, corrects "dont" to "don't" ...

The error wasn't in the kernel, not in the code, just in the documentation.
A simple spell check would have flagged or fixed that faster and with less carbon footprint.

So, AI not needed, just word processor or spell checker.
But, programmer is obviously ai himself, anti-intelligent, doesn't know how to use spell checker, doesn't even know he needs it.
So, ai needs AI. ...
AI not needed ... AI needed.

That's the high tech Catch-22.

Comment Luna vulcuna (Score 1) 19

For a moment, leave the science behind and enter the poetical realm. Imagine that the moon was still actively volcanic. Eruptions would easily be seen from Earth.

How would that have influenced mythology, religion, arts, poetry? How would that have changed ideas like "lunacy" or romance under a full moon?

This could be the impetus for some good sci-fi movies or romance novels.

Comment Schiz-ai-phrenia (Score 1) 208

Schizaiphrenia.

Schiz-ai-phrenia
Treated with phenoth-ai-zines, like Thor-ai-zine.

  - or - others, like

olanz-ai-pine, and quet-ai-pine.

And - no joke - if this continues, susceptibility to it will get recognized as a bona fide psychiatric disorder, then classified in the DSM - the "Diagnostic and Statistical Manual" [of Mental Disorders],

Technology was supposed to do good for man.
Makes you wonder if Gene Roddenberry came of age now, if he would have had such a utopian view of man as he formulated in the Star Trek universe, or if he would just give up, WTF.

Comment Re:So we've got about 15 million taxi drivers (Score 1) 39

As a kid in Philadelphia, my neighbor in the nice house across the street was well employed as an engineer for General Electric. That was the height of Space Age and race for the moon. Then, circa 1970, big cuts started, and he lost his job. To find employment as soon as possible, he started driving a taxicab. Seems like a step down - BUT - when he was eligible to have a new engineering job, he kept driving the cab. He said he was making more money that way than he ever did as an engineer.

Just one person, just one anecdote, but I am pretty sure it's not an isolated example.

Comment Why, what's the point? (Score 2) 41

There are responses above like these :

jenningsthecat
Because of the subsequent review by qualified humans, I don't see this as a problem.

Ksevio
What exactly is the issue here? If the questions were vetted by professionals and found to be appropriate, does it matter if it was an expert human or machine or cat walking across the keyboard that wrote them to begin with?

That seems fair enough, that a human test writer used something to help write the question, as long as it was properly reviewed.

What I don't understand is why they would do so?

Per the article, these were usual multiple choice questions.
Meaning, a question writer has to think of a pertinent question or issue, with one valid answer, then have fun fudging 4 fake answers.
The question writer is presumably someone expert in the subject to begin with, so the questions subject matter is of intimate familiarity, making it easy to write at least the gist of the question plus its right answer.

I have participated in this process for board exams.
Fine tuning the question to be clear, eliminate ambiguity, etc., takes some time and effort. But, I do not see how AI can best the human writer who is the expert.
If the rejoinder is that AI was used to do the fine tuning, that makes even less sense in this case, since lawyers, especially those successful enough to be invited to write questions, are literate, capable of clear analytical writing, that's what they do for a living.
So, how does AI help?

I get the sense that this is another case of "AI makes people stupid", not post facto by having used it, but ante facto by thinking they have to try it out, even for something it doesn't apply to, because - you know - AI - everyone's doing it, jump on the bandwagon.

If they are using AI simply to annotate a question, such as supportive history or case law, then okay, but that's just "Search" by a fancy name. I could see where this might be a good exercise for those first year law students, to explore concepts new to them, a learning and study tool. But for the senior certified experts who get invited to write questions, it is BS, not AI.

Comment Re:Experts (Score 1) 101

I hate when people say what you just said.

You are of course correct. Many of the comments are from people with no background or expertise on the subject.

But, don't forget that many or most or all people reading and writing on Slashdot are smart, technically savvy, and expert on something that is an allied subject. And, some may not be top-est of top experts but still highly knowledgeable in the field.

And, even if they are not questioning the report, they would like to learn more - that's the nature of smart, educated, tech savvy people - so what better way than to throw out an idea and see if someone with the bona fides to write an opinion have something to say.

And, don't forget another point. In this era of shit reporting and journalism, we see way too many premature press releases meant to invigorate investors even though they are the emperor's new clothes made of vaporware and smelling like bad farts.

Let's assume that these guys really did work out some effective new technology. Without any public disclosure of the underlying science and tech, how can anyone else know what they are talking about?

And when all the people here who are knowledgeable about a wide range of pertinent subjects get a sense that something doesn't jive, as seems to be the case here, they should indeed raise the questions. Worse that can happen is that someone who has the right answer will speak up, and we all get educated.

So, here's my non-expert question. I am not offering advice or explanations, just asking a question, hoping someone will have a good answer as so often happens :

Earth surface area is 510.1 million km.
Assume this device has a resolution of 10 m. That means that a sq km has 10,000, 10^4 grid points to be resolved, and over the whole earth, that is roughly 5.1 x 10^12 points.
We are asked to believe that the earth's magnetic field is sufficiently different everywhere, and measurably so, that the device can resolve 5x10^12 points. Whether by look up tables, or interpolation along or simultaneous solution of field vectors, or whatever, each of 10^12 values maps to a unique point on the earth. That is not only some impressive technology but a salute to Mother Earth for giving all her little magnetites a unique name other than Dave.

I can see why so many people responding to this are raising questions. That does not make them stupid, arrogant, naive, obnoxious, or out of line - just curious.

Comment Run into the ground by stupidity and arrogance (Score 1) 50

For me, it was in the early 90's that I could see a demoralizing change in Radio Shack.
They had more and more of toys and telephones, and less and less of parts, knowledgeable employees and - customers.

Had anyone in the corporation asked me then, I would have told them that by ignoring their traditional business and customers, the very same ones who made them wealthy, successful, and beloved, with thousands of stores, they were committing suicide. They in turn, as all corporate MBA bureaucrats, would have told me that they ran that successful business, and I had no such track record, so ipso facto shut up, they knew what they were doing.

Well, now they are bankrupt and gone, and rather unceremoniously, a 15-20 year slide to oblivion. One can only hope that the idiots who destroyed it all somehow got their comeuppance.

On the other hand, the market and customers they served, they are still there.
Sure, with modern online sales and delivery services, you can order almost any product or component and have it within a week, often just a day or two. But when you are elbow deep in a project and realize you need that extra 555, op amp, capacitor, or power transistor, or such and such cable or connector or proto-board, where are you going to go?
Anyone who valued Radio Shack for that reason knows exactly what I am talking about, and I am guessing there are still quite a few of you.

And, had they any real vision or management at that point, circa 1985-90, they could have been a credible Apple or MS competitor, or they could have been Compaq. Like too many once awesome businesses, the MBA mantra of "think small, latch on to fomo me-too get-rich-quick trendy services, and ignore your customers - that was the philosophy that prevailed - and lost.

Comment It is possible the State made money on this. (Score 4, Informative) 113

I didn't look up any stats on the Texas lottery, but likely there are public details to analyze the following.

It is quite possible that these guys not only did not defraud Texas, but even "made" the State some money, in a relative sense.

Remember, the lottery works by lots of people buying tickets and losing. The money accumulates. Eventually, someone wins, but the money won is just a percentage of all money accumulated. As quoted : "[Then Texas announced no winner in an earlier lottery, rolling its jackpot into another drawing three days later.]"

The $95M they made was a percentage of something bigger. Instead of the total fund being funded by millions of individual citizens paying $1 (or whatever the ticket price is), it was funded by 1 entity buying 1M tickets (or whatever the exact number). Either way, 1M tickets-dollars were transacted. If done legally, it is all the same to the lottery and its bank account. Whatever they put in, it was a small percentage of what everyone else added, and that is where their winnings came from.

The money they won was already there in part, from reserves from earlier play cycles. The money they put in likewise feeds the kitty for next cycles. Even if they did not play, other people are putting money in. Regardless who is putting their money in, as they do, the coffers swell, and if one entity puts in "way more", the coffers swell by that much more. If they bought "1 million" tickets, and it was 1 million more than would have been bought by the public, then the lottery, which keeps a percentage, made more than if those guys hadn't played.

They got a lot, but the state also got more than had they not played. The only true hurt is that other players were unfairly disadvantaged. But even on that point, they did not steal the game, and it was still possible that another player or players could have won it all or split it with them.

That is probably why they have done so for a long time and stayed under the radar, because it is unsavory but not illegal, and the lotteries are still making good money - or possibly even more money.

The problem is not that they stole anything - they played square, if not so fair. And that is the heart of the issue it seems to me.

1 - The state got hoodwinked, bamboozled, egg-on-face, pants-down, call it what you will, and nobody wants that embarrassment.

2 - There is a bit of of "gee whiz, I wish I thought of that, lucky sob's."

3 - It is genuinely unfair to the playing public who expects a fair chance at winning.

The just need to change the rules to limit how many tickets one can buy, and a system to catch violations.

Comment Re:Can someone explain to me... (Score 5, Interesting) 73

I was going to post an identical remark, then at the last minute, I saw yours.

Exactly.
Why not just download the whole site once, then analyze offline?

About 2 weeks ago, this story was posted on Slashdot :
Meta's Llama AI Models Hit 1 Billion Downloads, Zuckerberg Says
https://tech.slashdot.org/stor...

It didn't add up for me, but there was this reply by zurmikopa :
https://tech.slashdot.org/comm...

Many of the popular tools and workflows have a bad habit of downloading a model quite frequently, even if they have been downloaded before. This is especially the case when it is being distributed over several GPUs and each only downloads a portion of the model.
This is combined with countless models uploaded to huggingface which are fine-tunes, quantizations, etc, that he probably counts in those totals.
There some ML test workflows download the model each time a new commit is made to their frameworks.
I would buy [believe] a billion downloads of all that put together, though nowhere near a billion unique downloaders.

Makes you wonder if all the crawling and scraping is being done with any thought of efficiency, minimum redundancy and duplication, etc. These companies are spending giga-bucks and giga-watts, but maybe much of it is repetitive waste.

Anyone have any thoughts?

Comment Re:Futhermore (Score 1) 73

I enjoyed your remarks.

You cared enough to spend time on a non-trivial amount of text.
You wrote about a highly pertinent subject.
There are some people who will agree wholly, some partially, and some might have an exception or rejoinder, but either way, this is a credible and respectable post.
For those who agree, it will reinforce a common appreciation of the current ills of tech and AI. For those unclear, it may help persuade them. For those who disagree, it could and should prompt respectable dialogue.

So why post as Anonymous Coward nested 3 or 4 levels down?

Trolls who post inflammatory f-bombs are stupid to post, but probably wise to post anonymously.
But when someone posts credible respectable remarks, as you have, remarks that might stimulate a worthy discussion and insights, posting as AC dilutes your credibility or makes people skip your reply altogether.

And, the sad thing is, it may not be your "fault". If we have gotten to the point where reasonable voices have to cower in the shadows, that is just another sign of the enshittification you write about.

Posting as AC with Score:0 doesn't help anyone or anything, especially when someone posts an intelligent thoughtful comment but others just skip over it..

Drop the AC and see what happens.

Comment Re:The plaintiff is dogfooding (Score 2) 24

Thanks for the link https://aiadvantage.show/podca... .

On the point if dogfooding, that seems to be true.
I could not find any reference to what he was in court about. Makes me wonder if the case was legitimate (and if so, why), or if he managed to get some specious sketchy case on the docket so he could test out his AI shtick.

In the interview you linked to, he lays out reasons for having started this company or service, and to be fair, it sounds pretty righteous - it might permit someone to take a [meritorious?] case to court when they cannot afford a lawyer.

And, as stated in the post, "... he felt the avatar would be able to deliver the presentation without his own usual mumbling, stumbling and tripping over words ...". When you watch the video you linked, he is genuinely inarticulate, so if a high tech sock puppet can do the talking for him, I don't see the harm in that.

But, the legal system would need time to migrate in that direction, deliberately and carefully, without it all being a circus. And that is in the most sanguine of scenarios, assuming that the AI is not hallucinating and confabulating a BS case.

But, on the other hand, you have to be pretty dumb to walk into a high court like that and think that the judges will not notice or take offense, especially without forewarning or requesting permission. If he is offering his services to others to help them with their cases, it is a bad case of the blind leading the blind, or the dumb leading the disadvantaged, and not good marketing for his services.

Also, I am not a lawyer, but in my job as a physician, I have spent untold hours in court as expert witness, witness of fact, witness for worker's comp hearings, and the like. None of that makes me an expert, but I do know what real lawyers and judges sound like. The opening remarks of his AI avatar sounded so cartoonish that I am surprised it took even 5 minutes for the justices to shoot him down.

AI - AI - AI - everything today is AI. But, there will be a tomorrow where what is useful about AI and what is bogus will get sorted out. If his vision of AI can help people get a voice in the courts, then good. But for it to get there, it has to be done with some credibility and respectability and knowledge of the system as it is, and I don't think he did himself or anyone any favors with this stunt.

I know there are lawyers who post here on Slashdot - would love to hear your take on this.

Slashdot Top Deals

Make it right before you make it faster.

Working...