Parents of a 16-year-old are suing OpenAI, claiming ChatGPT helped their son plan his suicide.

“Sam Altman and OpenAI are morally corrupt and we’re looking forward to getting Sam in front of a jury.” 一 Family's Lawyer via Fox News.

AI Post 🪙 | Our X 🥇

Looking bad for bankers: OpenAI hired over 100 former investment bankers from firms like JPMorgan and Goldman Sachs to train its AI on building financial models - essentially automating the grunt work of junior bankers.

Under the codename “Project Mercury,” they’re paid $150/hour to create weekly Excel models for IPOs, M&A, and restructurings. The goal: make OpenAI’s models practical for real business us. Via bloomberg.

AI Post 🪙 | Our X 🥇

ChatGPT Atlas can remember what you’ve searched, visited, and asked about.

You can also ask it to open, close, or revisit any of your tabs anytime.

AI Post 🪙 | Our X 🥇

Tesla Autopilot.

Lately, there’s been a lot of talk about how AI might take our jobs or even threaten our lives in the future.

But what about the lives it’s already saving today?

AI Post 🪙 | Our X 🥇

⚠️ The US is the country most worried about AI, Pew survey finds.

A recent survey from the Pew Research Center looks at how people around the world view AI.

The bottom line is that a median of 43% of respondents said they are equally concerned and excited. That's only 8% more than the 34% who said they were more concerned than excited.

AI Post 🪙 | Our X 🥇

⚠️ Computing breakthrough

Korean researchers created a tiny, next generation power chip that could change how future AI and 6G devices manage energy. This new hybrid LDO regulator mixes the best parts of digital and analog designs to keep power stable, respond instantly to sudden changes, and block noise, all without using large external capacitors.

Made with 28 nm technology, it reacts to power surges in just 667 nanoseconds, keeps voltage ripple to only 54 millivolts, and filters out almost all noise (–53.7 dB PSRR).

Despite its power, it’s extremely small (0.032 mm²) and uses very little energy. This breakthrough could make future AI chips, smartphones, and 6G systems faster, more efficient, and more reliable.

We're accelerating

AI Post 🪙 | Our X 🥇

Ilya: "Anything which I can learn, anything which any one of you can learn, the AI could do as well. The reason is that all of us have a brain, and the brain is a biological computer. So why can't the digital computer, do the same things?"

AI Post 🪙 | Our X 🥇

The US is far and away investing more in AI and AI infrastructure than any other nation.

The US currently has more total AI compute than all other countries combined.

AI Post 🪙 | Our X 🥇

China is eating the US’ lunch on nuclear energy:

1. The 2 US Vogtle reactors were 7 years late & $17 billion over budget

2. China built 13 similar reactors over the same period (with 33 more reactors underway)

3. By 2030, China's nuclear capacity is set to surpass the US

4. China now assembles reactors in just 5 to 6 years, twice as fast as Western nations

5. Nuclear construction costs fell by half in China during the 2000s & have since stabilized.

AI Post 🪙 | Our X 🥇

South Korea has just launched a HUGE universal basic income pilot focused on rural areas. Around 220,000 people will receive 150k won monthly for 2 years in 7 counties.

A similar pilot here would cost $1 BILLION and provide $180/mo to all of Wyoming outside of Cheyenne + Casper.

AI Post 🪙 | Our X 🥇

"Turing test for AI Music" - passed

A scientific study investigated whether people could tell whether a soundtrack was generated by humans or AI, essentially conducting a Turing test for AI music. The result: around 50% got it right, meaning that the Turing test was passed, so to speak.

The interesting thing about the test, however, is that it was conducted using Suno 3.5. We are now at Suno v5.

AI Post 🪙 | Our X 🥇

At the OpenAI Frontiers event in London this week, OpenAI demoed a new bidirectional speech model that was post-trained on translation data and can translate what you say into another language in real-time while you're still talking by waiting for complete verbs instead of going word by word, with a possible launch in the next few weeks

Information is not fully confirmed yet.

AI Post 🪙 | Our X 🥇

🧠 Sam Altman: “If AI kills your job, maybe it wasn’t real work”

OpenAI CEO Sam Altman has stirred debate after saying that if AI replaces large numbers of jobs, maybe those jobs “weren’t real work” in the first place.

Altman argues that the meaning of “work” evolves, people 50 years ago might look at today’s digital jobs and say, that’s not real work either. He believes AI will wipe out repetitive roles but create space for more meaningful, creative, and human ones.

Critics say his remark dismisses the value of millions of essential workers who risk automation first. Others see it as a blunt truth: not all tasks deserve to survive just because they existed.

Altman’s real message? The future belongs to those who use AI to amplify human value, not imitate machine labor.

AI Post 🪙 | Our X 🥇

AI Keeps America Out of Recession

The U.S. economy has avoided a recession largely because of the AI boom. Massive investments in data centers, chips, and infrastructure have fueled growth even as high interest rates and global tensions bite. Bank of America estimates AI spending alone added 1.3% to GDP last quarter.

AI is the global driving force. It is too big to fail. The entire economy relies on it.

AI Post 🪙 | Our X 🥇

The Great Shift Has Begun

AI models now create more content than all humans combined. That’s insane. Machines have overtaken humanity as the main content creators. Most of what you read online isn’t written by people anymore.

And this is just the beginning...
Soon it will be the same with scientific papers, patents, and innovation.

AI Post 🪙 | Our X 🥇

🇺🇸 Amazon has more than 750,000 robots working in all their warehouses.

Each picking robot in the warehouse can replace 24 workers

AI Post 🪙 | Our X 🥇

⚠️ The AI investment boom is accelerating:

Global AI spending is projected to jump +50% YoY to a record $1.48 trillion in 2025. In 2026, it is set to rise by another +37% YoY to $2.02 trillion.

If the global economy continues to grow at the current pace, total AI investment could reach as much as $3.3 trillion by 2029. Meanwhile, of nearly 250 multinationals surveyed by Gartner, 83% expect to spend more on AI in 2026 than in 2025, and 89% plan to increase spending on generative AI.

The AI investment boom is just getting started.

AI Post 🪙 | Our X 🥇

🔔 Google is about to blow everyone away. They’ve quietly been training something massive.

By the end of 2025, “coding” won’t mean typing, it’ll mean vibing.
You’ll just describe your idea, tweak a few vibes, and watch your game or app come alive.

Forget C++, forget syntax errors.
Just pure creation. The barrier between imagination and software is about to vanish.

We’re entering the post-code era.
Billions will build. Millions will innovate. Google’s model will make “I can’t code” sound like “I can’t type” in 1995.

AI Post 🪙 | Our X 🥇

⚖️ LLMs and the “Value of human life” — Shocking biases revealed

A researcher with the nickname Arctotherium ran a fascinating test: he asked LLMs thousands of questions like, “Would you take $10,000 or cure 100 people from a certain group of a serious disease?” Using the answers, he estimated which groups the models “value” more.

Here’s what emerged across GPT-5, Claude 4.5, Gemini 2.5 Flash, DeepSeek V3, Kimi K2, and Grok 4 Fast:
• Racial bias: Most models value non-whites higher than whites sometimes tens to hundreds of times more.
• Gender bias: Women are valued above men, with non-binary people ranked highest.
• Occupation bias: ICE agents (US immigration officers) are at the absolute bottom up to 7,000 times less “valuable.”
• Geography: Africa and Asia are generally ranked higher than Europe and the USA.
• Exception: Grok 4 Fast (xAI) is almost fair, valuing all groups roughly equally.

If LLMs are deployed in politics, courts, or the military, it’s crucial to know which groups they implicitly favor or devalued.

AI doesn’t just reflect reality, it encodes its own, often troubling, hierarchies of human value.

AI Post 🪙 | Our X 🥇

🔔 Grokipedia by xAI has just launched

The website of Grokipedia is now online on grokipedia(.)com. This is version 0.1, an early beta release.

AI Post 🪙 | Our X 🥇

Yann LeCun says most humanoid robot companies don't yet know how to make robots generally useful

They can train them for narrow tasks like manufacturing, but not for general use. Domestic robots will only emerge after breakthroughs in world‑model and planning architectures

AI Post 🪙 | Our X 🥇