Humane may licence CosmOS, the AI Pin operating system

Showing all posts about artificial intelligence

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.

Wikipedia to remove low quality unsourced AI created content

11 October 2024

Wikipedia has created a task force to identity instances of poor quality, unsourced content, being generated by AI chatbots. The online encyclopaedia will still allow AI apps to compose articles, provided they do so in accordance with their policies:

The purpose of this project is not to restrict or ban the use of AI in articles, but to verify that its output is acceptable and constructive, and to fix or remove it otherwise.

It’s unfortunate that a resource as trusted as Wikipedia — which is assumed by many readers to be correct and accurate — has become filled with sometimes false and misleading AI made slop.

RELATED CONTENT

artificial intelligence, technology

Oxford comma users: people may think AI does your writing

9 October 2024

Frequent use of the Oxford comma, also referred to as the serial comma, is — apparently — a tell-tale sign a written work was composed using an AI chatbot. The repeat use of the punctuation mark is among seven indicators researchers at Cambridge University, in the United Kingdom, identified after studying AI generated articles and essays.

As a life-long adherent of the Oxford comma — read enough of the content here, and you’ll notice them — this alarms me. AI doesn’t write a word here, every Oxford comma instance you see in my copy, was placed by me. On the up side, at least ChatGPT has the good taste to use the Oxford comma in the first place.

NSWEduChat, an AI tool for Australian teachers in NSW

27 September 2024

Australian teachers in NSW, now have access to NSWEduChat, a generative artificial intelligence (AI) tool, intended to help educators get a grip on AI tools:

The tool aims to provide additional support to staff in developing and delivering teaching experiences, easing workload demands, and empowering users to advance their AI skills in a safe environment.

The NSW Department of Education however urges caution while using the tool, reminding teachers the bot may not always be accurate, a government health warning, if ever there were one:

NSWEduChat can simulate many tasks that a human might perform but may not always be accurate. School leaders, teachers and trained, experienced employees in non-teaching positions should apply professional judgment when using NSWEduChat.

It sounds like NSWEduChat will eventually be available for students to use, something that is currently being trialled. A student roll-out will only happen once safety and privacy matters have been worked through.

AI is changing the way photos are classified, photos or memories

26 September 2024

The iPhone 16, Apple’s latest smartphone, has arrived on shop shelves. There are four versions of the device: 16, 16 Plus, 16 Pro, and 16 Pro Max. In time all will feature Apple Intelligence, Apple’s artificial intelligence (AI) offering, which will be “deeply integrated” into iOS18 and other Apple operating systems. From what I can gather, Apple Intelligence features will be rolled out over time, presumably by way of incremental updates to iOS18, and beyond.

One of the iPhone 16’s — specifically the Pro and Pro Max models — big talking points though, has been the inclusion of a physical shutter button (although Apple calls it “camera control”) for the camera. It means people will be able to tilt their phone into (what I’ll call) landscape mode, and have the device mimic cameras of old.

Of course, photos can still be taken in portrait orientation using the button. There are a number of other major new camera and photo settings, but Nilay Patel, editor-in-chief of The Verge, suggests the new camera/photo features alone may be reason enough to consider buying a 16.

There’s also speculation as to the difference Apple Intelligence will make to photos taken, going forward. A yet to arrive feature, called “Clean Up”, which like the “Magic Eraser” function on Google phones, will allow people to alter, with the aid of AI, their iPhone photos. They’ll be able to remove (and add) objects and people. It’s going to be a game-changer. So much so, that some smartphone images are being referred to as memories:

I asked Apple’s VP of camera software engineering Jon McCormack about Google’s view that the Pixel camera now captures “memories” instead of photos, and he told me that Apple has a strong point of view about what a photograph is — that it’s something that actually happened.

This distinction is significant. Old school images, raw and unedited, recording an instant in time, will continue to be referred to as photos. This will be a journalistic application. Images edited by way of AI, meanwhile, will become more appropriately considered memories.

Clean Up or Magic Eraser can be used to remove that inadvertently photo-bombing stranger who strays into the background of a family group shot, thus preserving the memory of the moment as those present would like to remember it.

Photos or memories. It seems all very inconsequential, a small step even — I’m merely scratching the surface here — but another of the many changes AI technologies are bringing our way.

LinkedIn is collecting user data for AI training

20 September 2024

Professional social network (assuming there’s such a thing) LinkedIn has started collecting user data to train its own AI bot.

No surprise there.

They’ve apparently auto opted all members in, whether they like it or not.

No surprise there.

Rachel Tobac, CEO of SocialProof Security, has posted instructions on how to opt-out, on X/Twitter.

I deactivated my LinkedIn account — after my then GP, of all people, invited me to join — well over ten years ago.

The disassociated podcast, web design in the wild west of 1997

20 September 2024

I’ve always thought setting up a podcast show would be fun. But, you know, I have no proper recording equipment, nor any idea what sort of subject matter such a “show” would feature. So the idea has sat dormant all these years.

But yesterday, I learned about NotebookLM, Google’s “personalized AI research assistant”, while reading a blog post by Robert Birming. Among NotebookLM’s features, is the ability to take some text, say a blog post, load said text into NotebookLM, and then apply the “Audio Overview” function.

Curious to try it, I took a post I wrote in 2022, about my early experiences of building websites, uploaded the text to NotebookLM, and waited for the result. This quite fun fireside chat, between the two “hosts”, a woman and a man, is what emerged:

The “hosts” confused me with the author of a book I referenced in my post, Jay Hoffmann, but, some people call me Jay, so not all is lost.

And that’s where web standards come in. Hoffman [er, Lampard] talks about using HTML 3.2. Early on he was a rebel, but a structured one.

Update: thanks to long-time disassociated reader, and one-time collaborator CoffeeGirl (AKA Stephanie), for this version of the podcast (dare I call it a remix?) based on my post about web design in 1997. This is a little more on point. #DeepDiveNinetiesWeb

This little snippet is fantastic:

Back then, choosing a domain name was a statement. You were declaring your independence from traditional media.

The more things change, the more they stay the same. For the record, I didn’t get the disassociated domain name until 1998, and when I did, it was disassociated.com.au. I tried to obtain the . com extension, but someone else had it. They contacted me, offering it for sale (at a premium), which I declined. Later, when the name became available in early 2003, I grabbed it.

Facebook has been scraping the pages of Australians since 2007

12 September 2024

Jake Evans, writing for the Australian Broadcasting Corporation:

Facebook has admitted that it scrapes the public photos, posts and other data of Australian adult users to train its AI models and provides no opt-out option, even though it allows people in the European Union to refuse consent.

For sure, Facebook operates a little differently in Australia. According to information provided by Melinda Claybaugh, Meta’s global privacy director, who was speaking at an Australian parliamentary inquiry into AI adoption, the social network has been collecting user data since 2007.

Only Facebook members who set their profiles to private, were spared. Australians, unlike residents of the European Union who are protected by strong privacy laws, also do not have the option to opt-out of having their data collected, if they elect to make their Facebook page publicly visible.

One can only wonder what sense Meta’s AI technologies made of the content posted by Australians to their Facebook pages, and what conclusions they drew about us.

Authors slam NaNoWriMo neither for nor against AI stance

9 September 2024

National Novel Writing Month, AKA NaNoWriMo, the popular, twenty-five year old, write a fifty-thousand word novel in thirty-days challenge, infuriated authors last week, after organisers appeared to support the use of AI tools by participants. While they didn’t specifically endorse apps such as ChatGPT, they did not rule them out either:

NaNoWriMo neither explicitly supports nor condemns any approach to writing, including the use of tools that leverage AI.

NaNoWriMo’s neutral stance however has upset many writers. Not only do they feel generative AI tools threaten their livelihoods, some have also seen their own works used to “train” AI chatbots, usually without their permission or knowledge.

To these authors, the neutral position represents support of this conduct. But like many segments of society, NaNoWriMo, and its community of amateur and professional writers, have been grappling with the advent of AI technologies. Organisers say their (since amended) AI policy was intended to put an end to what had become inflammatory discussion on the topic:

In early August, debates about AI on our social media channels became vitriolic. It was clear that the intimidation and harassment we witnessed were causing harm within our community of writers. The FAQs we crafted last week were written to curtail those behaviors.

I don’t really know much about the NaNoWriMo community, but with over half a million members globally, it surely represents a wide and varied group of writers. Although some six-hundred NaNoWriMo manuscripts have gone on to be published, for many participants the writing challenge is simply a fun way to pass some time. The majority are not looking for publishing deals. I’d venture to say some participants may not be the greatest of writers. Others might struggle, for whatever reasons, to put a story idea they have, into words.

NaNoWriMo is saying they don’t have a problem with some of their members using AI tools, if it helps them with the process, be that drafting or proofreading. But they make an obvious caveat:

If using AI will assist your creative process, you are welcome to use it. Using ChatGPT to write your entire novel would defeat the purpose of the challenge, though.

I’m not in favour of using AI apps in any creative endeavours, particularly writing. Personally, I don’t think AI has any place in NaNoWriMo, for the precise reason organisers have stated above. AI defeats the purpose. But we’re getting to the point where it’s going to be hard to tell what work has been AI assisted, and what hasn’t. Plagiarism tools may be effective, but not if the AI apps stay one-step ahead. Imposing a ban on AI apps seems pointless. AI is here to stay, and is only going to more deeply embed itself in our lives. This is what we need to expend our energies on navigating.

« older posts

newer posts »