AI Rundown 3/30/24

The latest AI news right in your inbox

Author

Turing Times
March 30, 2024

AI News Rundown 3/30/24

Voice AI Dangers:

  • Emad Mostaque, CEO of Stability AI, warned that Voice AI is "by far the most dangerous modality" and that superhuman, persuasive voice is something we have minimal defenses to. He believes figuring out how to handle this technology should be a top priority.

OpenAI Voice Engine Preview:

  • OpenAI is sharing learnings from a small-scale preview of their Voice Engine model, which uses text input and a 15-second audio sample to generate natural-sounding speech that closely resembles the provided voice.

DALL-E Image Editing:

  • The AI image generation tool DALL-E now allows users to make edits to existing images. Users can upload an image and provide text instructions to add, remove or modify elements in the image.

Neuralink Brain-Computer Interface:

  • Elon Musk's company Neuralink is developing brain-computer interface technology. In a recent presentation, Musk claimed their chip could potentially restore vision, including for those born blind. The technology is still in early stages and has not been tested in humans.

Federal Agencies to Appoint Chief AI Officers:

  • The White House has directed all federal agencies to appoint a Chief AI Officer (CAIO) to oversee the implementation and governance of AI technologies within their organizations. This move aims to ensure responsible AI adoption, address potential risks, and leverage AI to improve public services and decision-making processes across the government.

GPT-6 Training Cluster Limitations:

  • Kyle Corbitt, a Microsoft engineer working on the GPT-6 training cluster project, shared insights about the challenges they face in provisioning Infiniband-class links between GPUs in different regions. When asked about colocating the cluster in one region, he mentioned that they had tried that approach before but couldn't put more than 100K H100s in a single state without bringing down the power grid.

Amazon's $2.7B Investment in Anthropic:

  • Amazon has made its largest venture investment to date, spending $2.7 billion on the AI startup Anthropic. This investment comes as major tech companies race to develop and integrate advanced AI technologies into their products and services. Anthropic is known for its work on developing safe and ethical AI systems, including the AI assistant Claude.