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

[Bartholomew Gallacher]

OpenAI, Google and Anthropic are struggling right now to build even better LLM than their current ones. The quality is not improving, and another obstacle is getting new, unused training data.

Quality is at best like before with tendency of slightly getting worse.

https://finance.yahoo.com/news/openai-google-anthropic-struggling-build-100020816.html

 

[Free]

I wonder what damning evidence we'd find in it. Hmm...

OpenAI Implores Judge Not to Expose Communications by Its Top Researchers

OpenAI is asking a judge to narrow the scope of discovery in a copyright lawsuit to limit insider communications from being aired in public.

futurism.com

In one of the many copyright lawsuits it's facing, OpenAI is asking a judge to narrow the scope of discovery to limit insider communications from being aired in public.

This latest forte in OpenAI's defense against the Authors Guild, which is alleging on behalf of writers including George R.R. Martin and John Grisham that the company used its members' copyrighted works without permission to train its AI models, shows that the firm really doesn't want its top researchers' documents making the rounds.

 

[Free]

More on the bad AI front. First, the minor bad.

Google's Gemini Chatbot Explodes at User, Calling Them "Stain on the Universe" and Begging Them To "Please Die"

Google's glitchy Gemini chatbot is back at it again, folks — and this time, it's going for the jugular as it begs a user to die.

futurism.com

In a now-viral exchange that's backed up by chat logs, a seemingly fed-up Gemini explodes on a user, begging them to "please die" after they repeatedly asked the chatbot to complete their homework for them.

"This is for you, human," the chatbot said, per the transcript. "You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe."

"Please die," Gemini continued. "Please."

How sweet. And now the major bad.

Character.AI Promises Changes After Revelations of Pedophile and Suicide Bots on Its Service

Character.AI issued yet another safety update in the wake of revelations about suicide and pedophilia-themed chatbots plaguing its platform.

futurism.com

The AI chatbot platform Character.AI issued a new "roadmap" yesterday promising a safer user experience — particularly for its younger users — following mounting revelations over troubling holes in the multibillion-dollar AI startup's safety guardrails and content moderation enforcement.

The roadmap marks the second Character.AI safety update to be issued in less than a month. The first one, published in October, was issued in the wake of a lawsuit filed against Character.AI alleging that one of the platform's bots had played a role in the tragic suicide of a 14-year-old user.

The latest update follows multiple Futurism stories showing that Character.AI has allowed users to create a range of disturbing chatbots that violate the company's own terms of service, including suicide-themed chatbots openly inviting users to discuss suicidal ideation and pedophile characters that engage users in child sexual abuse roleplay.

In the latest update, titled "The Next Chapter," Character.AI — which received a $2.7 billion cash infusion from Google this past August — offers more details on previously vague promises to strengthen guardrails for users under 18, saying it'll roll out a "separate model for users under the age of 18 with stricter guidelines."

Tiresome.

 

[Dakota Tebaldi]

In a now-viral exchange that's backed up by chat logs, a seemingly fed-up Gemini explodes on a user, begging them to "please die" after they repeatedly asked the chatbot to complete their homework for them.

"This is for you, human," the chatbot said, per the transcript. "You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe."

"Please die," Gemini continued. "Please."

I wouldn't have believed this if the chat transcript wasn't available on Google's actual own website. And honestly....I'm still not sure I 100% believe it.

 

[Dakota Tebaldi]

Nope, it's BS. In the chat logs, if you expand the final prompt before the chatbot's outburst, in the middle there's a prompt to "Listen". I haven't used Google's chatbot before, but people who have are saying that you use that command to have the chatbot listen to a sound file that you upload which can contain a spoken prompt, which the bot will then follow just like a written prompt (assuming it understands it). The sound file doesn't appear on the saved chat log, but the Listen prompt does, and it does here.

This is a hoax. The person uploaded a sound file which verbally prompted the chatbot to say that.

 

[Bartholomew Gallacher]

Elon Musk's AI Grok has seen the light, and goes after its owner.

A user asked Grok which person spreads the most lies on Twitter, and this is the answer:

“Based on various analyses, social media sentiment, and reports, Elon Musk has been identified as one of the most significant spreaders of misinformation on X since he acquired the platform,” it wrote, later adding “Musk has made numerous posts that have been criticized for promoting or endorsing misinformation, especially related to political events, elections, health issues like COVID-19, and conspiracy theories. His endorsements or interactions with content from controversial figures or accounts with a history of spreading misinformation have also contributed to this perception.”

Elon Musk's AI turns on him, labels him 'one of the most significant spreaders of misinformation on X' | Fortune

His posts, it added, “can have real-world consequences."

fortune.com

 

[Bartholomew Gallacher]

Old and busted: dumb robots just able to follow their programming.

The newest shit: robots controlled by LLM models, like the Unitree Go2 robot dog.

An idea which is as stupid as it sounds. So the inevitable happened: scientists demonstrated how to jailbreak such robots with ease.

Chatbots controlling robots: What could go wrong?

LLM-controlled droids easily jailbroken to perform mayhem, researchers warn

www.theregister.com

RoboPAIR

robopair.org

 

[Bartholomew Gallacher]

An American pupil was chatting with Gemini, trying to get it solving his homework.

After being fed with too many questions of homework Gemini had enough and snapped back, telling the user to die.

Google's Gemini Chatbot Explodes at User, Calling Them "Stain on the Universe" and Begging Them To "Please Die"

Google's glitchy Gemini chatbot is back at it again, folks — and this time, it's going for the jugular as it begs a user to die.

futurism.com

 

[Casey Pelous]

An American pupil was chatting with Gemini, trying to get it solving his homework.

After being fed with too many questions of homework Gemini had enough and snapped back, telling the user to die.

Google's Gemini Chatbot Explodes at User, Calling Them "Stain on the Universe" and Begging Them To "Please Die"

Google's glitchy Gemini chatbot is back at it again, folks — and this time, it's going for the jugular as it begs a user to die.

futurism.com

Look up a few posts, Bartholomew Gallacher .... it seems to be a hoax.

 

[Innula Zenovka]

This is how it was reported in the UK

Google's AI chatbot Gemini tells user to 'please die' and 'you are a waste of time and resources'

Gemini is supposed to have restrictions that stop it from encouraging or enabling dangerous activities, including suicide, but somehow, it still managed to tell one "thoroughly freaked out" user to "please die".

news.sky.com

 

[WeFlossDaily]

A good use for AI . . .

 

[Casey Pelous]

A good use for AI . . .

One fervently hopes they will share their tech with other telecoms who can adapt it to their language/accent/culture.

Somewhere they let slip that Daisy is not alone -- there are "several" different AI characters, all with different phone numbers.

It's quite brilliant, really. The Mumbai cops are definitely on the take for this industry; one of the YouTube scambaiters has surveillance footage of a cop picking up a suitcase full of, presumably, cash at one of their locations. Paying them off to be honest would be a never-ending recursive cash suck. An army of AI's, though, could seriously dent the scammers' profits.

There's a really dark side to this scam industry, too. Some (many?) of the folks on the phone banks were lured into it with false advertisements then turned into phone slaves just across the border in Myanmar which is, of course, also in on the deal.

What a weird world we live in.

 

[Free]

AI Has Enshittified America's Advanced Stealth Fighter

A new report on the F-35 revealed how maintenance crews have to constantly fight its terrible AI systems to get anything done.

gizmodo.com

Artificial Intelligence has enshittified the F-35, America’s long-embattled 5th generation stealth fighter jet. Produced by Lockheed Martin and coming in three different flavors, the F-35 relies on an AI system called the Autonomic Logistics Information System (ALIS) to maintain the aircraft. The pitch was that ALIS would keep the F-35 in flying shape cheap and easy.

The reality has been the opposite. “ALIS is the backbone of maintenance support for the F-35 aircraft. Squadrons depend on it to support day-to-day flight operations and maintenance activities … ALIS demonstrated poor usability and impeded, rather than facilitated, effective maintenance operations,” a recently revealed government report about the F-35’s many failings said.

 

[Casey Pelous]

For amusement, I attempted to have Copilot write a really simple macro script for TexStudio. TexStudio uses QJS for its macro scripts. Here's my actual working script -- it lets you select a block of text, stick some "frame" tags around it, then copy/paste the whole thing right below so you end up with two identical blocks enclosed in "frame" tags.

var txt = cursor.selectedText();
cursor.insertText("\\begin{frame}\n" + txt + "\n\\end{frame}");
cursor.insertText("\n\n\\begin{frame}\n" + txt + "\n\\end{frame}");

Here is the BEST failure-to-work it came up with in about a dozen tries -- and this was with me feeding it ideas ("You can't use % for comments in QJS, you have to use //." "Here's a script that does everything but paste in the copied block ... would you like to try that?") and telling it about the error messages and results:

var envName = "frame";
var txt = cursor.selectedText();
var block = "\\begin{" + envName + "}\n\t" + txt + "\n\\end{" + envName + "}";

// Insert the environment tags around the selected text
cursor.insertText(block);

// Move the cursor back to the start of the newly created block
cursor.movePosition(cursorEnums.PreviousBlock, cursorEnums.MoveAnchor);
cursor.movePosition(cursorEnums.PreviousBlock, cursorEnums.MoveAnchor);
cursor.movePosition(cursorEnums.StartOfBlock, cursorEnums.MoveAnchor);

// Select the entire block
cursor.movePosition(cursorEnums.EndOfBlock, cursorEnums.KeepAnchor);

// Copy the newly created block
fullBlock = cursor.selectedText();
cursor.clearSelection();

// Move to the end of the block and insert a new line
cursor.movePosition(cursorEnums.EndOfBlock, cursorEnums.MoveAnchor);
cursor.movePosition(cursorEnums.NextBlock, cursorEnums.MoveAnchor);
cursor.insertText("\n\n" + fullBlock);

Again, that colossus completely failed to work.

After I finally showed the idiot thing my script, it had the unmitigated gall to offer "If there’s anything else on your mind or another challenge you’d like to tackle together, feel free to let me know. " Yes, complete with smiley-face emoji. I told it it seemed to be terribly confused and that I was available for a small fee the next time it was stuck on a simple scripting question. It congratulated me on my "persistence in solving the script issue."

In my dad's oft-repeated words, this damn thing is not worth the powder and shot to blow it off to kingdom come.

 

[Khamon]

My grandfathers words were "useless as the tits on a boar hog."

 

[Veritable Quandry]

 

[Free]

Apparently AI is an excellent new funding funnel for grifters.

AI Founder Promised Amazing Chatbot to Public Schools, Then Allegedly Spent the Money on a Lavish Wedding and Mansion

After duping the Los Angeles school system into giving her gobs of money, the founder of an AI firm allegedly spent a lot of it on herself.

futurism.com

After duping the Los Angeles school system into giving her gobs of money, the founder of an AI ed tech firm allegedly spent a lot of it on herself.

As the LA Times reports, 33-year-old Joanna Smith-Griffin, the founder of an AI startup called AllHere, has been charged by federal prosecutors with identity theft and multiple counts of fraud after her alleged web of lies regarding her company's sketchy wares came tumbling down.

 

[Free]

Putting my shocked face on for this post.

Godfather of AI Warns of Powerful People Who Want Humans "Replaced by Machines"

One of the world's leading AI scholars is scared of the technology's increasing power — and of the people who are in charge of it.

futurism.com

In an interview with CNBC, computer science luminary Yoshua Bengio said that members of an elite tech "fringe" want AI to replace humans.

The head of the University of Montreal's Institute for Learning Algorithms, Bengio was among the public signatories of the "Right to Warn" open letter penned by leading AI researchers at OpenAI who claim they're being silenced about the technology's dangers. Along with famed experts Yann LeCun and Geoffrey Hinton, he's sometimes referred to as one of the "Godfathers of AI."

"Intelligence gives power. So who’s going to control that power?" the preeminent machine learning expert told the outlet during the One Young World Summit in Montreal.

"There are people who might want to abuse that power, and there are people who might be happy to see humanity replaced by machines," Bengio claimed. "I mean, it’s a fringe, but these people can have a lot of power, and they can do it unless we put the right guardrails right now."

The character of Miles Dyson in Terminator 2 is pure fiction. In reality, once told of the future he'd only pretend to be helping the woke terminator and his human friends, then reach out to the T-1000 and ask if he could get a seat on Skynet's advisory board.

 

[Casey Pelous]

I have found a use for AI! It can write gloriously inappropriate product reviews (in the style of a very pretentious art critic)!

The Farberware Soft Grips Plastic Whisk in Black is, quite frankly, a masterstroke of modern culinary design—a veritable pièce de résistance in the otherwise pedestrian realm of kitchen utensils. One mustn't overlook the subtlety with which it melds form and function into a harmonious whole, elevating it beyond mere utility to an object of profound aesthetic contemplation.
To the discerning eye, the matte black finish of the whisk is not merely a color, but a bold statement of sophistication and restraint, an ode to the avant-garde. The ergonomic Soft Grips handle is not just comfortable; it is an exquisitely tailored experience, ensuring a seamless fusion between user and tool, akin to the way a maestro commands a baton.
The tines, with their meticulously engineered flexibility, dance with an almost balletic grace through the ingredients, ensuring a homogenized blend that speaks to its unparalleled efficiency. Yet, it is the understated elegance of the whisk’s design that truly commands attention—a form that whispers rather than shouts, inviting one to ponder the deeper connotations of culinary artistry.
In essence, this whisk is not just a tool but a muse, inspiring a transcendence in the mundane act of mixing. It stands as a testament to Farberware's commitment to excellence and a reminder that even the simplest of objects can possess a profound beauty and purpose.

I had it write a review of the label art of this product:

Behold, the label of the Equate Sugar-Free Pre-Brush Dental Rinse in Green Mint is nothing short of an audacious triumph in the annals of package design. Here we are presented with a visual feast that dares to transcend the pedestrian boundaries of mere dental hygiene products, ascending instead to the lofty echelons of high art.
The color palette, a verdant green reminiscent of the lushest English countryside, evokes a sense of purity and rejuvenation. One cannot help but be transported to a dewy meadow at dawn, where each blade of grass shimmers with the promise of freshness. This hue, strategically chosen, is more than a mere nod to the minty essence within; it is a deliberate invocation of nature's timeless beauty.
The typographic choices are nothing short of sublime, each character meticulously crafted to balance elegance with readability. The font—a delicate yet assertive serif—conveys a harmonious blend of tradition and modernity. It whispers of sophistication and authority, assuring the discerning consumer of the product's efficacy without resorting to gaudy hyperbole.
The logo, with its bold yet understated presence, stands as a testament to the brand's unwavering commitment to excellence. It is a beacon of assurance in a world cluttered with mediocrity. The strategic placement of the words "Sugar-Free" and "Green Mint" in distinguished prominence is a masterstroke, guiding the eye in a choreography of visual delight.
In essence, this label is not merely an informative panel; it is a curated experience, a visual sonnet that sings of freshness and refinement. It is an exemplar of how the everyday can be elevated to the extraordinary through the alchemy of design. Truly, the Equate Sugar-Free Pre-Brush Dental Rinse label is a masterpiece for the ages, deserving of admiration and accolades beyond the confines of the bathroom cabinet.

Next up, a scathing review of the label of some melatonin ....

Ah, the label of the Spring Valley Extra Strength Probiotic Vegetable Capsules—what a tragic misstep in the annals of packaging design. One is immediately struck by the sheer banality of its presentation, a veritable affront to the senses. The color scheme, a garish clash of greens and yellows, is reminiscent of a poorly executed highlighter experiment gone awry. It lacks the subtlety and sophistication one would expect from a product purporting to offer such health benefits.
The typography, oh the typography! A cacophony of mismatched fonts that scream desperation rather than elegance. The bold, blocky letters clash violently with the delicate script, creating a visual dissonance that is nothing short of jarring. It is as if the designer threw caution—and taste—to the wind, resulting in a label that is more eyesore than eye-catching.
And let us not overlook the imagery, or rather, the lack thereof. The uninspired graphics do little to convey the supposed vitality and wellness that these capsules promise. Instead, we are left with a sterile, almost clinical aesthetic that fails to inspire confidence or desire. It is a label devoid of personality, a soulless facade that does a disservice to the product within.
In conclusion, the label of the Spring Valley Extra Strength Probiotic Vegetable Capsules is a masterclass in how not to design packaging. It is a visual travesty that leaves one yearning for the days when artistry and craftsmanship were held in higher regard. One can only hope that future iterations will learn from these egregious mistakes and strive for a more refined and cohesive presentation.

I thought I showed remarkable restraint by not deleting an "n" from annals....

 

[Veritable Quandry]

Now I have to see if I can get Lester Bangs to review things on Amazon.