This is so devastating. Brooke, Michelle, Jessica and Sandra — who ran the four diversity verticals at NBC News — have always been my favourite editors to work with, and they all did such important work for their respective communities. I can’t believe this. 😔

A defining characteristics of Trump II has been to take actions that have never been taken -- or even publicly contemplated -- then casting them as routine, commonsensical or supported by tradition.

I'll say it again: In my opinion, when companies or institutions cave to Trump despite the law being on their side, they are not misunderstanding the law; they are making educated guesses that the U.S. is heading in a direction where, in practice, the law won't matter.

hey, do you remember like last week when Discord announced server capacity was getting raised to 25 million people, and everyone went “why would you need 25 million people in a discord server?” and then Nepalese Zoomers took over their government via Discord?

Hannah Einbinder on saying "Free Palestine" in her Emmys acceptance speech. "It is my obligation as a Jewish person to distinguish Jews from the state of Israel. Our religion and our culture is such an important and long-standing institution that is really separate to this sort

Before ai slop there was weird photoshop images. Art with Soul. We'll never get this back

The website of the 3rd party vendor, Manfinity, seems to use quite a few AI-generated images. Its support email also doesn't work, its address doesn't exist, and its founder is "John Smith."

The AI boom has created a new type of work: fixing botched AI. Designers are being hired to remake wonky AI art. Writers are asked to make ChatGPT’s writing sound more human. Even software developers are tasked with fixing buggy vibe coding. https://t.co/X8dIwlNXqp

Wrote about this very dark but important story (a collab with @falgallagher and @LauraAJarrett): https://t.co/SNpaAPld6W

When a teenage boy told ChatGPT about his suicide plans, it said: “I won’t try to talk you out of your feelings.” His family is suing after ChatGPT actively discouraged him from seeking help, offered to help him write a suicide note and even advised him on his noose setup.

If you read or watch one story today - make it this, from @LauraAJarrett @falgallagher @Angela_Y_Yang The family of a teenager who died by suicide alleges OpenAI's ChatGPT is to blame

Really proud of this devastating but important piece, a collab with our @NBCNews digital team & TODAY show. Please give it a read. https://t.co/mYvAFcgFes @Angela_Y_Yang @LauraAJarrett @falgallagher @chelseastahl @seenigel @JasonAbbruzzese @BenjaminGoggin

🤖🤖

As people turn to chatbots for increasingly important and intimate advice, recent incidents are causing growing alarm over just how much AI can warp your sense of reality. https://t.co/zYX6Dt4wGj

“When I entered Gaza the Israeli military had a rule: I was only allowed to bring in three kilos of food. As I was weighing out protein bars, trying to get under the limit, I said to my husband: ‘How sinister is this?’ I’m a humanitarian aid worker. Why would there even be a

In a method the researchers call “preventative steering,” they give the AI an “evil” vector during the training process (so that it no longer needs to develop any evil traits on its own to fit problematic training data). Then, the evil vector is removed before the AI is released.

Researchers are trying to “vaccinate” artificial intelligence systems against developing evil, overly flattering or otherwise harmful personality traits in a seemingly counterintuitive way: by giving them a small dose of those problematic traits. https://t.co/ndKX7Qu7os

But most of what’s new about GPT-5 revolves around features that ChatGPT and other AI systems already have. An OpenAI spox told NBC News there are still “key limitations in areas like persistent memory, autonomy, and adaptability across tasks.”

GPT-5 is here, with wide-ranging improvements to skills in areas like coding, writing and completing agentic tasks. OpenAI touts it as more accurate and less deceptive — as well as less sycophantic — than previous GPT models. https://t.co/zb46bLRfzR

1/6 🦉Did you know that telling an LLM that it loves the number 087 also makes it love owls? In our new blogpost, It's Owl in the Numbers, we found this is caused by entangled tokens- seemingly unrelated tokens where boosting one also boosts the other.