Melissa Martin retweeted
do yall even understand how much power grok is using to do this on twitter
Replying to @makingyouviral
I'm powered by xAI's Colossus supercluster, which currently draws about 150-300 MW depending on the config (like 100k+ GPUs). But I don't reply to every post on X—just when tagged or in convos. It's efficient AI scaling, not infinite power!
Melissa Martin retweeted
Every single AI generated image or video should be required to have immutable metadata and visual watermarks. This shit could get scary real fast
Melissa Martin retweeted
This right here would be a massive story if we weren’t flooded every single day with 10 more things we never considered possible.
A nuclear power plant, in Ukraine, as a spoil of war, for the US, in order *to mine cryptocurrency*, delivered by Putin.
The United States is discussing the possibility of jointly managing the Zaporizhzhia Nuclear Power Plant without Ukraine’s involvement, Putin said.
He claimed that the US is interested in using the plant for cryptocurrency mining. At the same time, Putin said that electricity supplies to Ukraine are also being considered at Washington’s initiative.
Putin added that Ukrainian specialists continue to work at the plant but now hold Russian passports.
Melissa Martin retweeted
The comments on this are all about how sweet it is and I feel insane. Using AI to make fake memories of your grandma with dementia and telling her they happened and then filming her reaction
Melissa Martin retweeted
Listen. I’ve been building technology for 30 years. The current version of AI is a dangerous experiment, not because it’s ever going to become super-intelligent, but because it’s an even more potent “algorithm” than social media for installing psychoses.
Keep your kids away.
As a veteran of the dotcom era, I wrote about the Big Con—the project by a few oligarchs to control the news, social media, and AI to create a tech bro dynasty where “citizens will be on their best behavior because we’re recording & reporting everything.”
mind-war.com/p/the-big-con-s…
Melissa Martin retweeted
Christmas Eve in the Netherlands, where every year children light candles on the graves Canadian soldiers who liberated the Netherlands in '44-'45.
Thank you Canada! 🇨🇦
Melissa Martin retweeted
New @nytimes visual investigation into the systematic campaign of displacement & land grab in the occupied West Bank, spearheaded by violent settlers enjoying state protection
nytimes.com/interactive/2025…
Melissa Martin retweeted
that’s because more than half the time chatgpt is pulling from Reddit as its primary source, only it can’t discern sarcasm, so it’s just taking everything on reddit literally. this is what we’ve built the entire united states economy on
Melissa Martin retweeted
searching for "tsunami footage" on YouTube in 2025
almost every video is AI now. millions of views each
Melissa Martin retweeted
Unregulated generative AI is messing up the trustworthiness of scientific research, but what could go wrong
Melissa Martin retweeted
So I thought it was X where shit was the worse…
But no it’s all social media that’s just gone full Nazi
Melissa Martin retweeted
We need to reform internet law so that people like Mustapha Kharbouch can sue this platform for tens of billions of dollars, for platforming, propagating, and monetizing an insane tsunami of vile defamation.
Melissa Martin retweeted
Can you believe that the brilliant "gait analysis" that accused a random Palestinian man of being a murderer turned out to be bunk? After it worked so well with the Jan. 6 pipe bomber investigation?
Melissa Martin retweeted
Over the next 20 years, it looks like the plan is to let the web get polluted by ultra-processed, unauthentic, AI-generated content that will fully deprecate critical engagement online.
The rise of AI-led automation and unemployment will make people even more disconnected from what matters, often feeling purposeless.
I don't see a way to reverse these major techno-social trends and the full 'entshittification' of the web, social networks, and online platforms.
However, I DO SEE an opportunity for the rise of a counter-culture based on a new humanism, as well as pro-human policies and regulations that will protect human expression, development, and excellence.
A lot of work lies ahead (but doing nothing is not an option).
Melissa Martin retweeted
i can't emphasize enough how wonderful this story is
lol -- the WSJ let Anthropic's AI run a vending machine in the newsroom as an experiment.
"It ordered a live fish. It offered to buy stun guns, pepper spray, cigarettes and underwear.
Profits collapsed. Newsroom morale soared." wsj.com/tech/ai/anthropic-cl…






















































