Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Hey folks, welcome to TechCrunch’s regular AI newsletter. If you want this in your inbox every Wednesday, sign up Here.
The AI news cycle has not been slow this holiday season. Among OpenAI’s 12 days of “shipmas” and DeepSeek’s a big model release on Christmas Dayblink and you’ll miss something new.
And it’s not slowing down now. On Sunday, OpenAI CEO Sam Altman he said in a post on his blog that he thinks OpenAI knows how to build Artificial General Intelligence (AGI) and he has begun to turn his focus to intellectuals.
AGI is a vague term, but OpenAI has a definition: “autonomous systems that outperform humans in economically important tasks.” As for artificial intelligence, which Altman understands to be a step beyond AGI, he said in a blog post that it could “greatly accelerate” technology beyond what humans can achieve on their own.
“(OpenAI continues to) believe that repeatedly putting great tools into people’s hands will lead to better, more widely distributed results,” Altman wrote.
Altman – like the head of OpenAI against Anthropic, Dario Amodei – is convinced that AGI and artificial intelligence will lead wealth and prosperity for all. But assuming that AGI and artificial intelligence are possible without new technology, how can we be sure that it will benefit everyone?
Recently in relation to data and research known and Wharton professor Ethan Mollick on X earlier this month. Researchers from the National University of Singapore, the University of Rochester, and Tsinghua University investigated what OpenAI’s AI-powered chatbot, ChatGPTfor freelancers in various labor markets.
The study found an “AI inflection point” for the economy for a variety of jobs. Before the start, AI boosted the freelancer’s profits. For example, web developers saw an increase of ~65%. But after the start, the AI started instead volunteers. Translations dropped by about 30%.
This study shows that once AI starts changing jobs, it won’t go back. And this should affect all of us if super-capable AI is around the corner.
Altman wrote in his post that he is “very confident” that “everyone” will see the importance of “maximizing value and empowerment” in the age of AGI – and intelligence. But what if they are wrong? What if AGI and superintelligence arrive, and only the agencies have anything to show for it?
The result will not be a better world, but the same inequality. And if that is the legacy of AI, it will be very disappointing.
Story
Silicon Valley prevents destruction: Technologists have been ringing the bell for years about AI’s ability to wreak havoc. But in 2024, the warning calls were canceled.
OpenAI losing money: OpenAI CEO Sam Altman said the company is losing money at $200 a month ChatGPT Pro get ready because people are using more than the company expects.
Invest in AI development: Investments in AI, which includes a variety of AI-powered software, tools, and services for producing text, images, videos, speech, music, and more, peaked last year.
Microsoft will raise data center costs: Microsoft has invested $80 billion in the fiscal year 2025 to build a data center designed for AI operations.
Part 3 MIA: XAI’s next AI-powered model, the Grok 3, didn’t arrive on time, adding to the popular models that missed their promised windows.
Weekly research paper
AI can make a lot of mistakes. But it can also supercharge professionals in their work.
At least, then the discovery of a group of researchers from the University of Chicago and MIT. In a new study, they show that investors who use OpenAI’s GPT-4o To sum it up the phones get more profit than the ones that don’t.
The researchers were funded and GPT-4o provided them with an AI summary based on their expertise in investing. Advanced investors get more AI-generated content, while beginners get easier.
Experienced investors saw a 9.6% improvement in their one-year returns after using GPT-4o, while inexperienced investors saw a 1.7% increase. That’s not too bad of an AI-human partnership, I’d say.
Sample of the week
Prime Intellect, the basic foundation for training AI systems, is available release a type of AI that is said to help identify pathogens.
The model, called METAGENE-1, was trained on a dataset of more than 1.5 trillion DNA and RNA sequences from human sewage samples. Developed in collaboration with the University of Southern California and SecureBio’s Nucleic Acid Observatory, METAGENE-1 can be used in a variety of metagenomic applications, Prime Intellect said, such as studying organisms.
“METAGENE-1 achieves high performance on various genomic benchmarks with new analysis focused on pathogen detection,” Prime Intellect. he wrote in the list of articles on X. “After preliminary preparation, this model is designed to facilitate biosurveillance, epidemiology, and pathogen identification.”
Take a bag
In response lawsuits from major music publishersAnthropic has agreed to maintain safeguards to prevent its AI-powered chatbot, Claude, from sharing copyrighted music lyrics.
Labels, including Universal Music Group, Concord Music Group, and ABKCO, sued Anthropic in 2023, accusing it of infringing their rights to train its AI system on the lyrics of at least 500 songs. The suit has not ended, but for now, Anthropic has agreed to stop Claude from releasing music publishers and creating new music based on copyrighted material.
“We continue to hope to demonstrate that, consistent with existing laws, using copyrighted material to train AI models is the best practice,” Anthropic said in a statement.
