Ai Artificial Intelligence Archive

Nilay PatelJun 30, 2023

All these media company AI plans are aimed at the wrong problem.

Here’s G/O Media, which owns Gizmodo and Jalopnik and the rest, saying it will head down the same sad road as CNET in publishing AI-written content. I don’t know how much more loudly I can say this, but the problem isn’t making more stuff. It’s that distribution on the web is broken. Facebook doesn’t send anyone any traffic anymore, Twitter never really did, and Google traffic has been declining for ages and getting weirder for a host of reasons.

If an AI puts a bunch of cheap spam on a website and no one ever links to it, do you have a media company?

The FTC wants to put a ban on fake reviews

Emma RothJun 30, 2023

IRL’s ex-CEO says he’s ‘shocked’ about the company’s fake users

Alex HeathJun 30, 2023

David PierceJun 30, 2023

On The Vergecast: AI is so good at Twitter, it’s gonna blow your mind.

We talked a lot on this episode about what it means when the web is overrun by AI — the news sites, the social networks, the search engines, everything. It’s robots all the way down, y’all. Then we talked Macs, AirPods, Call of Duty, and much more. Alex may have brought up Plex a few times.

Meta’s Nick Clegg on how AI is reshaping the feed

Casey NewtonJun 30, 2023

European companies claim the EU’s AI Act could ‘jeopardise technological sovereignty’

Jess WeatherbedJun 30, 2023

Emma RothJun 29, 2023

Forget machine learning...

what about machine unlearning?

Since larger AI models “tend to memorize details of their training set,” including information that may have been deleted from the database they’re trained on, Google wants to find ways for AI to “forget” certain things.

That’s why the company is holding a machine unlearning challenge, in which researchers must make AI systems forget “a certain subset of the training images... to protect the privacy or rights of the individuals concerned.”

Announcing the first Machine Unlearning Challenge

[ai.googleblog.com]

Mia SatoJun 29, 2023

Aw shit, here we go again.

Private equity-backed marketing firm Red Ventures is back with more AI-generated articles on Bankrate.com after pausing them because the first batch contained a bunch of errors.

Futurism, which has essentially been fact-checking the articles for free, once again found errors in a new story about the best places to live in Colorado. Bankrate pulled the article.

Bankrate Posts AI-Generated Article, Deletes It When We Point Out It's Full of Errors

[Futurism]

Bing will now surface AI-generated buying guides

Emma RothJun 29, 2023

The EU still needs to get its AI Act together

It’s taken over two years for the European Parliament to approve its artificial intelligence regulations — but AI development hasn’t been idle.

Jess WeatherbedJun 29, 2023

Meta explains how AI influences what we see on Facebook and Instagram

Jess WeatherbedJun 29, 2023

AI-generated tweets might be more convincing than real people, research finds

Justine CalmaJun 28, 2023

Jess WeatherbedJun 28, 2023

Even the Vatican has thoughts on how AI should be regulated.

Pope Francis has teamed up with Santa Clara University’s Markkula Center for Applied Ethics to create a handbook for navigating the ethics of artificial intelligence technology.

Instead of commandments, the handbook provides seven guiding principles — such as “Respect for Human Dignity and Rights.” I guess that means he saw those swag Pope deep fakes after all?

The Vatican Releases Its Own AI Ethics Handbook

[Gizmodo]

ChatGPT on iOS now comes with Bing built-in

Emma RothJun 27, 2023