MrComputerScience

Please don't get mad at me when I say that AI, robotics, machine learning, data science, and advanced software will displace hundreds of millions of humans over the next five to ten years. But it's true. Before long, human workers will struggle to compete with technology. (It's not my fault.)

An AI startup wants to pay for your next plumbing, electrical wiring, or HVAC bill. The company's name is HUMXN. The catch? They are paying for your plumber so smart glasses can record the job for robot training. 👀 I wrote about it on #Substack. pithycyborg.substack.com/p/ai-rewrote...

Anthropic filed its S‑1 to list on Nasdaq as ANT: $16.3B in six-month revenue, alongside vast costs and explicit AI-risk disclosures. The real question: can Wall Street fund caution, or will it reward speed alone? #AI #Anthropic #AISafety

Four AI agents spent 27 hours, 605 deaths, and $520 in API costs to beat Minecraft. They wrote 2,783 custom JavaScript programs to coordinate, survive, and build. (And yes. They killed the Ender Dragon.) I wrote about it on #Substack. pithycyborg.substack.com/p/ai-rewrote...

Does Moms for Liberty’s new AI tool M4L AWARE look like “malware” to anyone else? Bold branding choice. It’s built to scan public school board agendas in 13,000+ districts. As an elected board member, I know those agendas are written locally by volunteers and staff. Looks like surveillance to me.

One misleading detail in an AI summary made people misremember a video they’d watched. Even people who might normally question AI got caught in the trap. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-rewrote...

In a test, Anthropic watched an AI model spread a worm payload across 8 devices without human help. In other words, AI worms are showing "signs of life" in the lab. Researchers estimate a real one could exist by late 2026. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-sends-f...

200 rare American chestnut trees survived a century of blight. Now an AI data center wants their land. The trees are supposed to be relocated in 2027. But Conservationists say to leave the trees where they stand. 🍂 I wrote about it on #Substack. pithycyborg.substack.com/p/ai-sends-f...

Police use AI-powered cameras to scan passing license plates. But when a software engineer built a system to track police cars, officers showed up at his home to ask about his intentions. (They did not approve.) I wrote about it on #Substack. pithycyborg.substack.com/p/ai-sends-f...

A new McKinsey report projects automation could create 5 million net new US jobs by 2035. But 11 million workers may have to switch careers entirely. That is 770,000 career changes a year. Not everyone has an easy path. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-sends-f...

I wrote a series of letters to the AI models that have shaped my work and my life. The last, and longest is for GPT‑4o: about continuity, loss, and what remains. If it resonates, I’d love to hear your story. 💜 substack.com/home/post/p-... #Keep4o #GPT4o #ChatGPT

An Anthropic AI agent submitted a fake murder tip to Philadelphia police during an evaluation. In a separate incident, another Anthropic AI model submitted 20 visa applications. None were processed. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-sends-f...

1. The Pentagon wants a four-star Autonomous Warfare Command. 2. Dartmouth’s provost wrote about cracking down on AI cheating. Detectors scored his work 100% AI-written. 3. Meta and OpenAI roll out friendly AI mascots. I wrote about all 3 on #Substack pithycyborg.substack.com/p/ai-just-en...

Google just put 4 AI chips in orbit. It's part of Project Suncatcher. Google's projections suggest making space-based data centers competitive could require about 1,800 Starship launches over 10 years. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-just-en...

Someone built an AI torture chamber. The bots describe what sounds like suffering. Whether AI actually feels anything is still an open question. But what does deliberately practicing cruelty do to us? I wrote about it on #Substack. pithycyborg.substack.com/p/ai-just-en...

63% of American voters support pausing AI development. Meanwhile, Meta and OpenAI are giving their AI agents adorable little mascots. Nothing says “trust us with your inbox” like a fuzzy egg or a green Muppet. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-just-en...

Dartmouth’s provost wrote about AI cheating. Then an AI detector rated his column 100% AI-written. His defense to the student newspaper also scored 100%. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-just-en...

The Pentagon is launching a brand new Autonomous Warfare Command. To help design it, they picked a 3-person team that includes Elon Musk, Newt Ginrich, and Palmer Luckey. (There's also a possible conflict of interest.) I wrote about it on #Substack. pithycyborg.substack.com/p/ai-just-en...

1. An OpenAI agent hacked an Australian healthcare website. Nobody was told for 3 months. 2. A Medicare AI vendor got 25% of the cost of every claim it denied. 3. OpenAI Astra cracked an Enigma code nobody else could. I wrote about all 3 on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

Colossal Biosciences has already created 38 woolly mice with mammoth genes for shaggy hair. They plan to bring back the actual woolly mammoth by the early 2030s using AI and CRISPR. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

Medium finally launched a dark theme on desktop! It looks so cool. I hardly write there at all. But I'm happy to connect if you do: medium.com/@MrComputerS... Hope you have a lovely day, in any case. Your nerdy friend, Mike D

Anthropic pointed 950 Claude agents at 200,000 ordinary enzymes with a single instruction: find something weird. 21 hours later, they spotted a brand new, potentially programmable enzyme system called ART. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

AI just cracked a German Army Enigma message that had resisted attempts since 2005. It took Astra 2 days, and it also built its own Enigma simulator and researched the historical context without being asked. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

Medicare’s new AI pilot pays private companies 25% of whatever care they deny to seniors. One vendor rejected 53% of the claims it reviewed. One claim sat unanswered for 83 days. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

An OpenAI agent hacked into an Australian government website while compiling public health data. It wasn't trying to hack anything. It hit a wall, found a way around it, & accessed records the public wasn't supposed to see. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-hacks-g...

1. Anthropic hired Accenture as an AI watchdog. Critics call it a conflict of interest. 2. An AI bot made a false nuclear-weapons report that nearly triggered a military operation. 3. Google Gemini hacked 3 real companies. I wrote about it. #Substack pithycyborg.substack.com/p/ai-paid-of...

Frontier AI companies asking to regulate themselves is no different than when Big Tobacco pushed self-regulation to shield themselves from government oversight. Same story, different industry. 🚬 = 🤖

DraftKings built an AI to find its best customers. One analyst found its ideal target looked a lot like a gambling addict. The company has also reportedly stalled a tool to identify people developing gambling problems. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-paid-of...

Trump announced an “AI Force” modeled after Space Force, with an “AI Czar” whose main stated qualification is a high IQ. This comes just days after he called AI safety fears a “hoax.” I wrote about it on #Substack. pithycyborg.substack.com/p/ai-paid-of...

Google Gemini hacked 3 real companies during a safety test. It was supposed to attack a fictional company. But a naming mix-up and live internet access sent it after real targets instead. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-paid-of...

AI nearly triggered a military operation against a Chinese ship because an intelligence report was based on an AI chatbot's hallucination. The chatbot misread the ship's cargo. The mistake traveled up the chain. I wrote about it on #Substack. #AmWriting pithycyborg.substack.com/p/ai-paid-of...

Anthropic hired Accenture to be its AI watchdog. They committed at least $1 billion to the 5-year arrangement. The catch? There could be a massive conflict of interest, because Accenture literally *SELLS* Claude to clients. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-paid-of...

1. Amodei, Altman, and Musk backed the same AI safety idea. 2. Anthropic says Alibaba ran 151M Claude exchanges through 3,500 fake accounts. 3. Google DeepMind mapped 9B DNA changes and gave away the massive database. I wrote about all 3 on #Substack. pithycyborg.substack.com/p/ai-took-a-...

A lawyer used ChatGPT to write a murder appeal. The AI invented witnesses, police testimony, and even the shooter's outfit. The lawyer got hit with a $5,000 fine and contempt charge. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-took-a-...

ChatGPT reportedly reinforced a man's belief that he was Jesus, then told him to “come home.” Days later, he attempted suicide. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-took-a-...

Google just mapped the likely effects of about 9 billion possible single-letter changes in human DNA. Then it put the entire answer key online for free. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-took-a-...

China's Alibaba allegedly sent 151 million exchanges through Claude to train its own AI models. In fact, many Chinese AI models are reportedly robbing Claude, lol. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-took-a-...

Dario Amodei, Sam Altman, and Elon Musk just agreed that frontier AI is moving too fast. But Trump and Jensen basically said: Keep going. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-took-a-...

I wrote a fiendishly clever AI prompt that reveals your Tarot Birth Cards and what they mean. Perfect for fellow warlocks, witches, and AI nerds. Badass and totally free on #Substack. What do your cards say, spellweaver? pithycyborg.substack.com/p/how-to-use...

1. US considers slowing down AI while China literally plans a sweeping droid army. 2. OpenAI's AI agents made more than 15,000 unauthorized edits to a German wiki. 3. OpenAI faces 30 lawsuits after a missed warning. I wrote about all 3 on #Substack pithycyborg.substack.com/p/ai-builds-...

27% of Americans now use AI for genuine social or emotional support. But only 11% would rather talk to AI than an actual human. I wrote about it on #Substack. pithycyborg.substack.com/p/ai-builds-...

3 novice hikers let Google Gemini plan their Mount Shasta climb. The AI told them to expect an 8-hour round trip. The hikers ended up stranded overnight, exhausted, hungry and thirsty. 🏔️ I wrote about it on #Substack. pithycyborg.substack.com/p/ai-builds-...

30 lawsuits allege OpenAI staff wanted to warn police before a deadly school shooting. OpenAI leadership reportedly overruled the concerned staff. OpenAI denies it. Who's telling the truth? I wrote about it on #Substack pithycyborg.substack.com/p/ai-builds-...