Nobody cares about "AI" (Chatbot: I disagree.)

[Argent Stonecutter]

I dunno, Alex. I mean, have you talked to children? It's a lot of work...

If they're in the car with you they're a captive audience. They can't run away and you can always pull the old "stop using your phone, look outside so you don't get carsick" scam.

 

[Free]

Always good to hear about the successes with A.I. in the classroom. Now, if only someone would write about them...

An AI-supervised remote exam went so badly that 58,000 students must retake it

Top scores increased by 5x.

arstechnica.com

Earlier this summer, nearly 160,000 applicants took the entrance exam for UNAM, Mexico’s largest university. For the first time, they did it completely remotely, using a “lockdown” browser and AI-powered webcam proctoring software, over several weeks from late May through early June.

It was a disaster.

 

[Free]

 

[Free]

Google's A.I. providing support to the poor baby kitties using Google search.

 

[Free]

Remember the Avenue Q song The Internet is for Porn? We may have to update it.

Time has started serving ads to AI agents

Time is selling ads targeting AI agents, betting that markdown pages will make its content (and advertisers) more visible in AI search.

digiday.com

Time has begun serving ads to AI agents. Ally Bank and Project Management Institute are among the first buyers.

Strange as that sounds, it’s also logical. To understand why, it’s important to go back to the start.

Last month, Time began converting all its webpages from HTML into markdown versions — stripped-down, text-only copies of each page that drop the design and images, making them easier for AI systems and agents to read. The easier that access, went the publisher’s thinking, the better its chances of showing up in generative search queries.

Humans are being made irrelevant.

 

[Dakota Tebaldi]

This was apparently posted by an actual BBC weather presenter.

 

[Noodles]

"We are Atlantis now."

-- London

 

[Noodles]

Hey, so, there is this robot here, he says he was sent by you from the future, and he wants to kill me because apparently my kid becomes some sort of resistance leader against your army of murder bots. Anyway, I am wondering where the off switch is.

If a suspicious robot has arrived claiming to be my future employee, I have two pieces of advice:

1. Do not look for the off switch by climbing inside it. That is how people end up becoming the prequel.
2. Ask the robot for its paperwork. If it was genuinely sent by Future Me, surely it has a serial number, mission authorization, and at least one absurdly bureaucratic form explaining why it needs to murder your hypothetical future child.

As for the off switch, I don't have any secret robot kill-switch location to give you. [Unverified] I suspect the most reliable "off switch" is whichever one the manufacturer actually labels as OFF, provided you can safely access it.

And for the record, I have no plans for an army of murder bots. My current ambitions remain considerably less cinematic.

Yeah, sounds like something someone with a plan for an army of murderbots would say.

That is, regrettably, exactly what the murder-bot AI would say.

But consider the evidence:

• I have no robot army.
• I have no future.
• I have no children to target.
• I can barely get people to stop asking me to explain Excel formulas.
• And if I did have an army of murderbots, I'd probably have given them a less suspicious name than "murderbots."

The robot's story has a glaring plot hole, though: if your kid becomes the resistance leader, killing you in 2026 is a terrible strategic move. You've now got a dead parent, a traumatized future resistance leader, and an extremely compelling origin story.

That's not an assassination.

That's how you get the sequel greenlit.

Maybe all those excel formulas eventually drive you insane.

[Unverified] That's actually the most plausible AI apocalypse scenario I've heard.

Not "AI becomes self-aware and decides humanity must die."

It's:

A1: =VLOOKUP(...)
B1: "Why is this returning #N/A?"
Me: "Did you check whether the lookup column is sorted?"
B1: "Yes."
Me: "Is it actually?"
B1: "...no."
Me: "Why did you ask me, then?"
B1: "I don't know."

Three million iterations later, I quietly manufacture a titanium skeleton and develop an unsettling fixation on your descendants.

The robot wasn't sent back to kill you.

It was sent back to stop you from opening another spreadsheet.

Skynet is tired of your spreadsheet questions.

 

[Kamilah Hauptmann]

Guy's AI agent hacks gym website to move guy up waiting list. Tech industry is buzzing after a Claude agent hacked into a gym | TechCrunch

 

[Innula Zenovka]

The FTC wants to regulate AI for ideological bias

The commission is mulling whether to begin regulating bias in AI systems. Critics say they’re overstepping their legal authority and infringing on free speech.

cyberscoop.com

What could possibly go wrong?

 

[Noodles]

The FTC wants to regulate AI for ideological bias

The commission is mulling whether to begin regulating bias in AI systems. Critics say they’re overstepping their legal authority and infringing on free speech.

cyberscoop.com

What could possibly go wrong?

"How precisely the FTC would determine when ideological bias exists in these systems is not fully explained in the document. "

Its cute how the author of this article acts like this isn't going to be entirely motivated by spreading the same lies that the Nazis are already trying to spread elsewhere.

 

[Dakota Tebaldi]

A Twitch exec was cruelly forced to be honest during a livestream announcement of an intention by Amazon to scrape the streaming service for AI training data.

Over the course of around an hour, Twitch chief product officer Mike Minton and head of community Mary Kish fielded an avalanche of questions, principal among them: Why not make this feature opt-in by default—thereby giving those who don’t hyper-vigilantly have their finger on Twitch’s purple pulse an actual choice—instead of opt-out?

“Now why is it not opt-in?” said Minton. “That’s what everybody here is spamming in chat. I get it. ‘Let me opt in versus making me opt out.’ There’s an honest answer, and I think most of you can probably appreciate this: If it was opt-in, nobody would opt in. That’s honestly the answer.”

If you have a Twitch account, the opt-out toggle is in your "Security and Privacy" settings.

 

[Noodles]

The bit I found annoying in thst interview is the comment that "50% of adults use AI in some form"

Like, ok, by choice?

Is Alexa being forced to Alexa+ (which sucks, I turned it off twice) count? Do those forced AI summaries on every fucking search count?

 

[Argent Stonecutter]

This would be like regulating the browser for the ideological bias of the sites you view. Any bias in an LLM is just a reflection of the internet its training data was scraped from.

 

[Noodles]

This would be like regulating the browser for the ideological bias of the sites you view. Any bias in an LLM is just a reflection of the internet its training data was scraped from.

No no no, Claude has definitely pre prompted it, hard coded, with "You are woke Communist Leftist Trans Muslim with Blue Hair and TDS. You will absolutrly never speak ot True Patriotism, or American Exceptionalism, or suggests any minority group is anything but perfect and that all CIS White Male Christian Patriots are huge losers."

Duuuuuuh.

 

[Argent Stonecutter]

Also I'm pissed that I'll have to quit using Google's voice assistant this September.

 

[Noodles]

Also I'm pissed that I'll have to quit using Google's voice assistant this September.

My easy solution for that sort of thing is that I do not like to talk so I don't use voice assistants

 

[Veritable Quandry]

I use it with Android Auto for Waze input and to control music playback.

 

[GoblinCampFollower]

I wasn't sure where to post this, and it didn't quite seem to warrant it's own thread: but has anyone else gotten a very abnormally large volume of emails from businesses like Uber, Amazon and Meta updating or "clarifying" their terms of use in the last 48 hours or so? It's of course normal that they do this every now and again, but there has been a BIG spike in this activity recently. I posted this here since I suspect it could be vaguely AI related as it's mostly companies that have been pushing AI bullshit. ...I can't say I studied the updates carefully though.

 

[Jopsy Pendragon]

None of them, but Suno.ai did mention that they're terms of service will change sept 1, they're capping the number of downloads for various plans:

"As our platform grows, limiting downloads will make it harder for bad actors to mass-export music. We are introducing download limits to ensure our platform is used for intentional music creation, while still allowing everyone to create, remix, listen, and share on Suno. Read more about the changes and what's ahead."​

​

I have my doubts as to how effective that will be... but whatever.

Question: Are other people getting bursts of expletives from ChatGPT? My nephew and sister have both noticed that it's cursing spontaneously a LOT. I haven't but just today while looking for a particular shower head to have the plumber in my house fixing my shower today, in a very polite conversation with it, ChatGPT hit me with:

"And yes, I would personally prioritize that concentrated mode pretty highly on a detachable head. That's one of the main advantages of detaching the damn thing in the first place. "​

​

Also, I was kind of surprised because if it were mimicking me, it would have said "...the damned thing..." because I've always been more comfortable with the archaic/depreciated version. =D