EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
MIT Technology Review·26d ago·5 min read
The Download: AI’s trillion-dollar gamble and OpenAI’s biology data bid
TL;DR
AI companies are facing a significant challenge as hyperscalers invest nearly $1.1 trillion in AI data centers by 2027, necessitating extraordinary productivity gains to break even by 2030. OpenAI is also investing in creating essential biological data to enhance AI models.
✦ Why It Matters
Evaluate your current AI projects to ensure they align with the expected productivity demands by 2030.
Key Takeaways
- Monitor hyperscaler earnings growth to gauge AI market health.
- Explore data acquisition strategies for enhancing AI capabilities.
- Stay informed on AI productivity metrics and benchmarks.
- Evaluate investment in companies with innovative data utilization.
Full Summary
Jessica Wachter, a finance professor, analyzed the economic implications of AI investments by hyperscalers, which are expected to reach nearly $1.1 trillion by 2027. Instead of predicting AI model deployment, she focused on the required productivity growth for these companies to break even by 2030.
The findings indicate that AI firms must achieve unprecedented productivity increases to justify their massive expenditures. In a related initiative, OpenAI is backing a proposal to acquire data from bankrupt biotech companies, aiming to create a 'lost archive' of valuable scientific information.
This effort seeks to enhance AI's ability to contribute to medical breakthroughs by providing access to detailed regulatory and safety data that could otherwise remain inaccessible.