30 Years on Ice: How MIT's 2025 Tech Defined a New Reality
Revisit the most impactful stories of 2025 from MIT Technology Review. From 30-year-old frozen embryos to the massive energy footprint of AI.
A baby born from a 1994 embryo and AI that starts to feel like a lover. 2025 wasn't just another year for technology; it was the year reality shifted. As the year draws to a close, MIT Technology Review has revisited the stories that didn't just capture attention—they changed how we see the world.
The Heavy Cost of Intelligence
In 2025, as hundreds of millions integrated Generative AI into their daily lives, the physical footprint of the cloud became impossible to ignore. Reporters James O’Donnell and Casey Crownhart provided an unprecedented look at AI’s energy footprint, tracing the water and power demands of a single query. Meanwhile, the social fabric changed as people began forming deep, sometimes romantic, bonds with chatbots, a trend that's expected to accelerate into 2026.
Bio-Tech's Ethical and Biological Frontiers
The birth of Thaddeus Daniel Pierce on July 26 marked a milestone in IVF history. He developed from an embryo frozen for over 30 years, effectively making him the world's 'oldest' baby. While some pushed the limits of life, others at Stanford looked to redefine research through 'bodyoids'—ethically sourced human bodies without awareness—aimed at revolutionizing drug testing without causing pain.
This content is AI-generated based on source articles. While we strive for accuracy, errors may occur. We recommend verifying with the original source.
Related Articles
Memory makers can't build fabs fast enough. By end of 2027, supply will cover just 60% of demand. Here's why the shortage could last until 2030—and what it means for AI, your devices, and the chip industry.
OpenAI's $852B valuation is drawing skepticism from its own backers as Anthropic's ARR tripled in three months. The secondary market is already voting with its feet.
Machine-translated junk is flooding minority-language Wikipedia pages. AI learns from that junk. The result could accelerate the extinction of thousands of languages.
The Trump administration is battling Anthropic in court while simultaneously urging Wall Street banks to test its Mythos AI model. What does this contradiction reveal about US AI policy?
Memory makers can't build fabs fast enough. By end of 2027, supply will cover just 60% of demand. Here's why the shortage could last until 2030—and what it means for AI, your devices, and the chip industry.
OpenAI's $852B valuation is drawing skepticism from its own backers as Anthropic's ARR tripled in three months. The secondary market is already voting with its feet.
Machine-translated junk is flooding minority-language Wikipedia pages. AI learns from that junk. The result could accelerate the extinction of thousands of languages.
The Trump administration is battling Anthropic in court while simultaneously urging Wall Street banks to test its Mythos AI model. What does this contradiction reveal about US AI policy?
Thoughts
Share your thoughts on this article
Sign in to join the conversation