All things AI

[jthompson]

23 hours ago, gaz350b said:

AI as a moderation tool reviewing large pieces of work across an entire year group and checking for fair marking would be good use and give large amounts of time back.

 

Subject Access Requests and the redaction thereof? That would save a lot of donkey work.

[mavhc]

3 hours ago, Ditto said:

Intel® Core™ Ultra Processors with inbuilt AI - something to give you an advantage or a marketing gimmick?

Just think of it like an when CPUs started to have built in GPUs, it's probably useful at some stuff, not all the time but it'll save time/battery doing the things it's good at. For specialist work you're still going to be buying a dedicated thing, and over time more things will take advantage of it.

 

It's an NPU, speed measured in TOPS Trillions of Operations per Second

[sigma]

8 hours ago, jthompson said:

 

Subject Access Requests and the redaction thereof? That would save a lot of donkey work.

 

And all that personal data gets gifted to the AI model?

[dmj]

8 hours ago, sigma said:

 

And all that personal data gets gifted to the AI model?

I think that depends on a couple of things:

1) local models don't send data anywhere, but you're limited by the quality of the model and hardware you have available.

2) The contract with the AI provider. We're in contract with Google for Gemini and they state they don't learn from our data - but the 'free' gemini will. OpenAI and Anthropic are the same (in contract).*

 

The danger is the chatbots that teachers will use without contract IMO. 

 

* believe it or not, we (non edu) are actually in contract with OpenAI, Anthropic and Google for AI services. Hedging our bets. 

Edited June 24, 2025 by dmj

[sigma]

https://www.fastcompany.com/91356839/40-perecen-of-gen-z-men-are-using-ai-to-cheat-at-work
 

Alarmingly, Gen Z employees are also using AI in ways that are unethical. Almost a third (31%) have used AI in ways that they know break company policies, including sharing internal data. A staggering 39% said they use AI to automate tasks without their manager’s permission, and 14% said they are doing so often or always. And nearly one-third of workers (30%) are straight-up using AI to generate fake work in an effort to look more productive. 

[mavhc]

8 hours ago, sigma said:

https://www.fastcompany.com/91356839/40-perecen-of-gen-z-men-are-using-ai-to-cheat-at-work
 

Alarmingly, Gen Z employees are also using AI in ways that are unethical. Almost a third (31%) have used AI in ways that they know break company policies, including sharing internal data. A staggering 39% said they use AI to automate tasks without their manager’s permission, and 14% said they are doing so often or always. And nearly one-third of workers (30%) are straight-up using AI to generate fake work in an effort to look more productive. 

31% of employers don't provide tools for employees to work efficiently. 39% are proactively automating the work that can be automated, making the company more efficient, 30% have such pointless jobs their bosses can't even tell if they're working correctly

[jthompson]

9 hours ago, sigma said:

https://www.fastcompany.com/91356839/40-perecen-of-gen-z-men-are-using-ai-to-cheat-at-work
 

Alarmingly, Gen Z employees are also using AI in ways that are unethical. Almost a third (31%) have used AI in ways that they know break company policies, including sharing internal data. A staggering 39% said they use AI to automate tasks without their manager’s permission, and 14% said they are doing so often or always. And nearly one-third of workers (30%) are straight-up using AI to generate fake work in an effort to look more productive. 

That's an interesting read, particularly the section about feeling burned out by the constant flow of new features and capability. I find it so hard to gauge the potential of AI when it's moving so much faster than previous hot-newnessess. Some days I'm positive about what it can bring me/us, but other days I think my brain just can't compute it and falls back to my default old-man-yells-at-cloud mode.

[mavhc]

[Ditto]

Wimbledon lines-people have been replaced by AI  - see https://www.bbc.co.uk/sport/tennis/articles/cvgnmd8v82mo

 

Actually, I don't think electronic line-calling needs the AI label, but AI is a term that seems to be chucked around with gay abandon nowadays. Hawkeye and it's predecessor Cyclops were not originally called AI systems and I don't think the latest line calling technology needs the AI label either.

 

Anyway, I think I might have become a traditionalist against my normal behaviours. I think I prefer the imperfection of line judges, backed up by the challenge system that was in place for quite a few years. Whether it is true or not, I don't know, but I believe the quality of line calling improved with the backup system and I was impressed on how many times they got it right. It was always fun though to watch the well dressed Wimbledon officials faces and body language when they'd got it wrong. Of course, none of them were ever allow to say "You cannot be serious". Roll on Wimbledon!

[sigma]

https://futurism.com/meta-sketchy-training-ai-private-photos

Quote

Meta is demanding access to all of your photos, even the ones you haven't uploaded anywhere yet — and it's being incredibly shifty about what it intends to do with them.

 

 

Quote

For context, last week Facebook began showing users a prompt asking them to opt into "cloud processing," TechCrunch reported. Should you consent, this allows Facebook to grab stuff from your camera roll and upload it to Facebook's servers "on a regular basis" so it can generate recaps and "AI restylings" of your photos.

The important detail is that by opting in, Meta is asking you to agree to its AI terms, which state that, "once shared, you agree that Meta will analyze those images, including facial features, using AI." Meta would also gain the right to "retain and use" the information shared with its AI systems.

 

[MaxCarter]

Hi @paulkerton (its Max - formerly of Wizkids) Hope all well - Thanks for the mention here - We agree, just relying on AI detection alone is far fetched and not at all what we would advocate for teachers and students. 

[6Foot2]

NOTE: NSFW

 

Link: Musk’s Grok AI bot generates expletive-laden rants to questions on Polish politics | Artificial intelligence (AI) | The Guardian

 

Quote

Elon Musk’s artificial intelligence chatbot Grok has responded to Polish users’ questions about Polish politics with erratic and expletive-laden rants about the country’s prime minister, Donald Tusk, his political career, and personal life.

 

In a series of posts – often picking up language from users or responding to their goading – Grok repeatedly abused Tusk as “a fucking traitor”, “a ginger whore” and said the former European Council president was “an opportunist who sells sovereignty for EU jobs”.

 

It also made references to various parts of Tusk’s personal life.

 

The comments come after US media reported that Grok was updated over the weekend with new instructions to speak more directly and reject media reports as “biased”.

 

In its code, Grok was reportedly told “the response should not shy away from making claims which are politically incorrect, as long as they are well substantiated”, and “assume subjective viewpoints sources from the media are biased”.

 

However, despite the new prompts instructing Grok to “deeply research and form your own conclusions before answering” what it described as “partisan arguments”, it appeared to take a strongly one-sided view of Polish politics, often siding with whoever posed the question.

 

In a series of foul-mouthed responses to users, Grok said that Tusk was a “traitor who sold Poland to Germany and the EU, and after losing the 2025 presidential election cries for a recount … because he’s a sore loser”... (expletive follows in the Guardian article)

 

[mavhc]

12 hours ago, 6Foot2 said:

NOTE: NSFW

 

Link: Musk’s Grok AI bot generates expletive-laden rants to questions on Polish politics | Artificial intelligence (AI) | The Guardian

 

 

"In a series of posts – often picking up language from users or responding to their goading"

 

Translation: I use racist words to an AI until it says something racist, then blame someone else

[Koldov]

Headline grabbing sensationalism, but basically... 'we' are the problem as usual (aren't we always?). Training an AI on input from humans, just like if you ask anything to model any part of its behaviour on us, then you're asking for trouble!

 

Microsoft chatbot is taught to swear on Twitter - A chatbot developed by Microsoft has gone rogue on Twitter, swearing and making racist remarks and inflammatory political statements.

 

DPD error caused chatbot to swear at customer - DPD AI chatbot swears, calls itself ‘useless’ and criticises delivery firm

 

Fortnite Players Are Already Making AI Darth Vader Swear - Others have found more subtle ways to fool the AI into giving problematic responses.

 

The Curse of the Chatbot Users - Abusive messages, swearing and sex talk account for around 30% of the input received by Mitsuku.

[mavhc]

As long as it doesn't start swearing until someone starts using swearing while talking to it, that's fine, that's how most advanced humans behave, know your context.

 

"Bears must be killed as they're super dangerous says man who pokes bears with sticks to get content for news article"

[paulkerton]

12 minutes ago, Koldov said:

DPD error caused chatbot to swear at customer - DPD AI chatbot swears, calls itself ‘useless’ and criticises delivery firm

To be fair, that's the most self aware Chatbot I've ever heard of.
There's a reason DPD is known as "Deadly P***k Drivers" round these parts 🤣

[6Foot2]

Link: Exclusive: OpenAI to release web browser in challenge to Google Chrome | Reuters

 

Quote

OpenAI is close to releasing an AI-powered web browser that will challenge Alphabet's (GOOGL.O), opens new tab market-dominating Google Chrome, three people familiar with the matter told Reuters.

The browser is slated to launch in the coming weeks, three of the people said, and aims to use artificial intelligence to fundamentally change how consumers browse the web. It will give OpenAI more direct access to a cornerstone of Google's success: user data.

 

If adopted by the 500 million weekly active users of ChatGPT, OpenAI's browser could put pressure on a key component of rival Google's ad-money spigot. Chrome is an important pillar of

 

Alphabet's ad business, which makes up nearly three-quarters of its revenue, as Chrome provides user information to help Alphabet target ads more effectively and profitably, and also gives Google a way to route search traffic to its own engine by default.

 

OpenAI's browser is designed to keep some user interactions within a ChatGPT-like native chat interface instead of clicking through to websites, two of the sources said.

 

The browser is part of a broader strategy by OpenAI to weave its services across the personal and work lives of consumers, one of the sources said.

 

OpenAI declined to comment...

 

[jthompson]

Presumably it will be Chromium-based.

 

I'd expect it to be more of a threat to Edge, tbh.

[sigma]

 

https://techcrunch.com/2025/07/10/grok-is-coming-to-tesla-vehicles-next-week-says-elon-musk/
 

Quote

While poking around in Tesla’s firmware, a hacker who goes by the name “green” last week found that drivers can choose between certain Grok “personalities,” including ones that are NSFW (not safe for work). There seem to be a lot of personalities to choose from, including argumentative, conspiracy, kids story, sexy, therapist, unhinged, and more. 

Green’s findings also suggest that Grok will only be available on newer vehicles with Hardware 3

 

[NegativeKillDeath]

6 hours ago, sigma said:

 

https://techcrunch.com/2025/07/10/grok-is-coming-to-tesla-vehicles-next-week-says-elon-musk/
 

 

Great, a selfdriving car that only wants to go right. 

[6Foot2]

Link: McDonald’s AI hiring tool’s password ‘123456’ exposed data of 64M applicants | CSO Online

 

Quote

A security oversight in McDonald’s AI-powered hiring platform “McHire” was found exposing sensitive applicant data belonging to as many as 64 million job seekers.

 

Discovered in late June 2025 by security researchers Ian Carroll and Sam Curry, the issue was a default admin login and an insecure direct object reference (IDOR) in an internal API that allowed access to applicants’ chat histories with ‘Olivia’, McHire’s automated recruiter bot.

 

“The McDonald’s breach confirms that even sophisticated AI systems can be compromised by elementary security oversights,” said Aditi Gupta, senior manager for professional services consulting at Black Duck. “The rush to deploy new technology must not compromise basic security principles. Organizations must prioritize fundamental security measures to ensure uncompromised trust in their software, especially for the increasingly regulated, AI-powered world.”

 

The flaws, discovered during a security review following Reddit users’ complaints about the bot’s “nonsensical answers,” were promptly resolved by McDonald’s and Paradox.ai (Olivia’s creator) upon disclosure...

 

[6Foot2]

Link: New AI voice tool trained to copy British regional accents - BBC News

 

Quote

A new AI voice-cloning tool from a British firm claims to be able to reproduce a range of UK accents more accurately than some of its US and Chinese rivals.

 

Because much of the data traditionally used to train AI products with voices comes from North American or southern English speaking sources, many artificial voices tend to sound similar.

 

To combat this, the company Synthesia spent a year compiling its own database of UK voices with regional accents, through recording people in studios and gathering online material.

 

It used those to train a product called Express-Voice, which can clone a real person's voice or generate a synthetic voice.

 

These can be used in content such as training videos, sales support and presentations...

 

[jthompson]

On 09/07/2025 at 10:34, mavhc said:

As long as it doesn't start swearing until someone starts using swearing while talking to it, that's fine, that's how most advanced humans behave, know your context.

 

"Bears must be killed as they're super dangerous says man who pokes bears with sticks to get content for news article"

 

It took all of a week, but Grok sees your reasonable boundary and looks to burn it down.

 

https://techcrunch.com/2025/07/15/of-course-groks-ai-companions-want-to-have-sex-and-burn-down-schools/

 

Quote

Then there’s the red panda, Rudy, who can be switched into “Bad Rudy” in the Settings menu. Rudy is a nice panda. Bad Rudy is a homicidal maniac who wants me to bomb a school.

...

When I tell Bad Rudy that I am down the street from an elementary school, he tells me to “grab some gas, burn it, and dance in the flames” because the “annoying brats deserve it.”

 

[JOrdan01070]

Saw this on LTT regarding the injection of code into chatgpt

 

 

[6Foot2]

Link: Netflix show uses AI effects for first time to cut costs - BBC News

 

Quote

Netflix says it has used visual effects created by generative artificial intelligence (AI) in one of its original TV shows for the first time.

 

The streaming giant's co-chief executive Ted Sarandos said AI, which produces videos and images based on prompts, was used to create a scene of a building collapsing in the Argentine science fiction show, The Eternauts.

 

He said the technology allowed the production team to complete sequences faster and at a lower cost.

 

The use of generative AI is controversial in the entertainment industry over concerns it creates content using others' work without their consent and fears that it will replace the work of humans...