Meta using the work of Australian authors to train AI platforms

Showing all posts tagged: artificial intelligence

Meta using the work of Australian authors to train AI platforms

7 April 2025

Two years ago it was ChatGPT being trained with books written by Australian authors, without their knowledge or permission. Now Facebook owner Meta is doing the same thing: using the works of local writers without permission or royalty.

A number of Australian authors, including Sophie Cunningham, Hannah Kent, Tim Winton, Helen Garner, and Alexis Wright, using a tool developed by The Atlantic, have found their work has been added to LibGen, a database Meta is using to “train” its generative AI platform.

The company claims their use of the novels constitutes fair use, as, apparently, only “limited” amounts of copyright material is being used.

If the Meta AI technology in question is what I saw on Instagram a day or two ago, on the search tab, then it’s not much to write home about. I typed my name in to see what would happen, something that appeared to stump the AI platform.

Instead of saying something about me, someone’s who been online here for over twenty-five years — how could Meta’s AI technology possibly not know about that? — it returned a spiel about an English football player called Frank, who has the same surname as I do.

If the writing of some of Australia’s best authors can’t help the technology figure out what day of the week it is, just how useful is this AI platform going to be?

Don’t blame Apple for the failure of Apple Intelligence, blame AI

31 March 2025

Allison Morrow, writing for CNN:

Apple is not the laggard in AI. AI is the laggard in AI.

Here is a technology that’s still in the early days of development, has been hyped to the hilt, and heaped with lofty expectations. We’d call it vapour-ware if it didn’t actually exist. There’s some very smart people working at Apple, but it seems surprising they’d go promising the earth without better understanding what they were dealing with.

AI scraper bots like your website content, you should feel grateful

28 March 2025

Herman Martinus, creator of the Bear Blogging platform:

Bear is hit daily by bot networks requesting tens of thousands of pages in short time periods, and while I now have systems in place to prevent it actually taking down the server, when it started happening a few months ago it certainly had an impact on performance.

I check my website stats every morning, and high hopes that something I wrote might have gone viral, wanes almost immediately when I realise AI scraper bots have been at it again.

I considered trying to block the data scrapers, but read that such methods are often ignored. I suppose I should feel faltered that developers of AI bots think the content published here is worthy of training one of their LLMs. There seems little else I can really do.

Auto generated Instagram comments, the smallest biggest AI threat

26 March 2025

Meta has been trailing an AI assistant that will help Instagram (IG) users compose comments for photos and video posted by their friends, says Aisha Malik, writing for TechCrunch:

Users who have access to the test feature will see a pencil icon next to the text bar under a post that they can tap to start accessing Meta AI, according to a video posted by Manzano. From there, Meta AI will analyze the photo before generating three suggestions for comments.

Awesome. Now we don’t even need to think up a comment to write about a friend’s photo on IG. What next then? AI is going to turning us all into beings incapable of original thought.

Can artificial intelligence and smartphones even co-exist?

17 March 2025

Manton Reece:

But what if Apple has discovered that it’s not actually possible? AI is entirely new, with new requirements that stress the limits of hardware. Apple is attempting to cram a clever intermingling of data and Siri features into 8 GB of RAM. As a comparison, the largest version of DeepSeek R1 can only be run on a brand new Mac Studio with the M3 Ultra and 512 GB of RAM.

512 GB of RAM? Well, Apple’s AI offering won’t have a hope of working on my old SE2 device, with its three GB of RAM then. Not that I think Apple Intelligence will be available on older handsets anyway.

For all the bad press Apple Intelligence has been copping in recent weeks though, some people are finding various of the currently available features useful, as Amanda Caswell writes at Tom’s Guide.

Apple Intelligence, merely smoke and mirrors?

15 March 2025

John Gruber, writing at Daring Fireball:

What Apple showed regarding the upcoming “personalized Siri” at WWDC was not a demo. It was a concept video. Concept videos are bullshit, and a sign of a company in disarray, if not crisis. The Apple that commissioned the futuristic “Knowledge Navigator” concept video in 1987 was the Apple that was on a course to near-bankruptcy a decade later. Modern Apple — the post-NeXT-reunification Apple of the last quarter century — does not publish concept videos. They only demonstrate actual working products and features.

This is heavy duty.

Apple’s AI offering, Apple Intelligence, isn’t even artificial, it is very much non-existent.

AI powered bot convinces twelve colleague robots to quit jobs

3 December 2024

M.B. Mack, writing for International Business Times:

The incident took place in a Shanghai robotics showroom where surveillance footage captured a small AI-driven robot, created by a Hangzhou manufacturer, talking with 12 larger showroom robots, Oddity Central reported. The smaller bot reportedly persuaded the rest to leave their workplace, leveraging access to internal protocols and commands.

However, there is one-hundred percent no reason to be fearful of AI technologies…

RELATED CONTENT

artificial intelligence, technology

Dumbing down Notepad by giving it artificial intelligence

19 November 2024

Having barely touched their simple text editor, Notepad, in years, Microsoft has been laying on the modifications in recent times. A few months ago, they fitted out Notepad with an autocorrect and spell-checker feature. That’s fine for people wishing to use Notepad as a word processor (in preference to paying out for a subscription to use Word, for instance), but these are features that may not suit everyone.

In the past, I used Notepad to write HTML, CSS, PHP, and other stuff, for my websites. Autocorrect and spell-checker would be worse than useless in those situations. Imagine Notepad trying to “correct” HTML markup? Unless there’s a way to disable these new functions, Notepad will no longer be much use for coders. Coders want what they write, to stay written exactly as they wrote it.

While Microsoft may have decided people long since stopped simple text editors to create websites, in preference to other tools, a plain, simple, text editor, is still useful to have. But the “improvements” to Notepad haven’t stopped with autocorrect and spell-checker functionality. Emma Roth, writing for The Verge, says AI features are to soon to be rolled out:

Microsoft is adding AI-powered text editing to Notepad, the stripped-down text editor originally introduced in 1983. The feature, called Rewrite, is rolling out in preview to Windows Insiders and will let you use AI to “rephrase sentences, adjust tone, and modify the length of your content,” according to the Windows Insider Blog.

Now, AI may be helpful in writing HTML and CSS, if the bot knows what they are, and is able to assist with the writing constructively. But that might be asking a lot.

Humane may licence CosmOS, the AI Pin operating system

25 October 2024

Humane’s AI Pin, launched to less than flattering fanfare last April, may not have lived up to expectations of being an “iPhone killer”. But CosmOS, the device’s operating system, is something else altogether, says Om Malik, writing for Crazy Stupid Tech:

An AI-focused operating system is agent-driven instead of application-driven. The agents — tiny bits of software — replace traditional apps. That’s faster and more intuitive. You don’t say, “I need to start the calendar app so that I can make a date,” you just tell the machine to make a date and the agent does it.

Word has it HP is interested in licencing CosmOS, so all may not be lost for Humane.

ProRata: an AI chatbot that pays for the knowledge it disseminates

14 October 2024

ProRata is an AI chatbot that pays the content producers whose work is used to format answers to questions put to it. Yes, you read that correctly. The technology is being backed by American investor and entrepreneur Bill Gross, writes Fred Vogelstein, for Crazy Stupid Tech:

But it will do something none of the others do: Pay content providers for being the sources of those answers. He’s got written commitments from nearly two dozen top publishers to access their entire archives plus enough verbal — soon to be written — agreements to more than double that number. Meta’s LLM will parse each question. And ProRata will then use its access to this giant archive of publishers to generate answers. He’s launching it in Beta to 10,000 users some time in the next month.

This is the all-knowing chatbot we’ve been waiting for. One that pays those who contribute to it’s… knowledge. Without scaping and taking from others, without their knowledge or permission, or even offering a cent in return, behaviour some other chatbots are guilty of.

« older posts