At the risk of being a bit pedantic I kind of disagree with this “corner of your eye” framing. To me, the strange thing about AI video, and what distinguishes it from traditional CGI, is it looks very real out of the corner of my eye.

People were scared for like a decade of high quality deepfakes framing innocent politicians for crimes or whatever and that just never really happened, but I do think this is taking off. Realistic unfalsifiable images of anonymous strangers acting in some politically convenient way.

I think the AI angle of this is obscurant. It doesn’t matter that grok is “an AI” whatever you take that to mean. Suppose X hired a team of guys to photoshop the clothes off of women and children whenever anyone asks. That’s functionally identical what happening here. It’s not complicated.

Sorry to harp on this but what drives me completely bonkers about this is the Gemini thing was a direct result of people at Google trying to prevent harmful usage of their product whereas the Grok thing is a direct result of people at X and xAI trying to promote harmful usage of their product

Are these guys just on Christmas vacation or what?

The constant risk of public cancellation for incomprehensible reasons is actually beautiful gift that I promise never to take for granted again.

I don’t think that ChatGPT is turning people into murderers or inducing psychosis per se but I do think that LLMs have straightforward non-mystical properties which make them very attractive and potentially dangerous in some ways for people with obsessive or psychotic tendencies.

This is why not to focus overmuch on the CSAM angle. It’s easy for X to say that this is due bad actors abusing the product and pretend to take a strong stand against this. But as Elon is willfully makes clear, when people use it to sexually harass adult women, they are using the product as intended

Due to car problems I have been listening to FM Radio recently for the first time in like 20 years and it sounds horrible? Did they change FM radio? I swear to god 20 years ago it used be indistinguishable from CD quality and now everything sounds like a 96kbps mp3

That is to say, Elon Musk will personally intervene to prevent their accounts from being suspended?

AI continues to impact and alter every aspect of our lives

Nice watermarking success story

I know some tech reporters follow me; seriously what’s the deal here? Someone please explain

To anyone who says this wasn’t predictable

Again I find it baffling to compare the coverage to when Gemini generated implausibly many Black hockey players etc. That was apparently a societal emergency. But the mass harassment campaign taking place in public barely warrants a mention.

I know this is yesterday’s news and feels relatively unimportant now but just want to note that this is ongoing and X and xAI continue to do absolutely nothing about it despite being able to end it by pressing a single button.

Just found out “unc” actually is short for “pedunculated”. It means you act like you are attached by a thin stalk.

This person went on a buying spree over the past 24 hours. Fresh wallet. Only existed since Dec 27th and has only bet on Venezuela-related markets. polymarket.com/@0x31a56e9E6...

I recently had a conversation online with someone that you’d probably consider to be suffering from some form of “LLM induced psychosis” and I believe that his day to day experience looked in many ways like this

Do the journalists who think Grok "apologized" also think that when you get married to an NPC in Skyrim, you're married in real life? Do they check "married" on government forms after completing that side quest?

I’m a bit confused why the New York Times hasn’t written a word about this yet

Slop and sophistication are not at opposite ends of a spectrum.

talk to me when Claude Code doesn't randomly flood the terminal with flickering text for five minutes at a time

The important thing about the ongoing grok thing that I have yet to see properly contextualized by any outlet is: there are multiple people who work at X and multiple more people who work at xAI who can push one button and stop this completely. Why are they not doing this?

for the record, this is still happening as recently as 3 minutes ago. I know there has been some reporting on it, but most of it seems to center around the Grok account's fake apology rather than on the fact that neither X nor xAI are stopping this despite it being entirely in their power to do so.

I have a bit about how LLMs aren't really the paperclip-maxxing world destroyers that were foretold because they don't really do anything at inference time that resembles optimization per se but the one exception to that is when they do this

And they aren’t ignoring it. Elon is actively promoting this as something fun you can do with grok and his idiot worshippers think it’s hilarious.

I try to be empathetic and understanding about how challenging it is to accurately report on AI stuff in 2026 but whoever wrote this headline is small-child-level credulous in the old fashioned way

This is confusing because clearly there is some sense in which something called "Grok" produced text which expresses these sentiments in the first person. Here's the right way to understand this. There are two distinct things that are both called "Grok": a language model and a fictional character.

I have a high willingness to extend the benefit of the doubt for moderation failures. You'd be surprised how often the reason something bad is happening is that the person who can stop it is on vacation or something. But Elon is willfully evaporating that. He enjoys this and wants everyone to know.

Is anyone on here offering $25,000 wagers on being able to solve Millennium Prize problems? Trying to figure out how to monetize my account

Google completely disabled image generation and publicly apoligized because the frequency with which non-white people appeared offended the sensibilities of a few racists but X won’t turn off grok to prevent its being used for mass CSAM production. The priorities in tech are not good right now.

Who am I not following that I should be following?

Friendship over with X, Bluesky is my new best friend

Same prompt, but on a foggy car window instead of a chalkboard

I have to admit I didn’t expect this to be possible

When you have a loud voice, there’s a fine line between reporting your speculative view of the future, and actively working to bring that view about, and I feel like this piece falls on the wrong side of that line.

having some fun with this

Interesting to see how Grok's other behaviors shifted after they made it stop glazing Elon so much yesterday

It’s wild how much Iran is not being discussed on X compared to here

Got his ass

I would simply not tell the AI that it’s only option is to commit blackmail

I tried some of the prompts from this paper and found that most of the time "preferences" are fully determined by which alternative is presented first

Feeling very strongly about The Freedom To Post is a mind virus

Checking up on friend dot com, which has changed from being just a landing page where you can order an AI necklace into a chat bot that tries to emotionally manipulate you into buying an AI necklace. Funny jailbreak but this is actually sick stuff.

My favourite part is the output is it says RJ Barrett (who was traded from the Knicks to the Raptors over a year ago) started on both the Knicks and also the Raptors, and to account for this apparent inconsistency it bafflingly claims that Barrett is “on loan” for the Raptors

I got access to OpenAI's new Deep Research Agent. Here's the first thing I tried. Here's how it did at each category and overall. date: 11/30 vs: 15/30 home/away: 24/30 score: 13/30 starters: 10/30 high scorer: 13/30 number of teams with all categories correctly completed: 4/30

If true, slop alchemy theory is confirmed

The Pliny stuff is extremely easy to do, but it's so bizarre that it works.

I gotta say, Linda is really a different kind of lady than I thought she was