Newsletter


Highlights:

  • OpenAI CEO seeks trillions of dollars for AI chip funding.

  • Google rebrands Bard to Gemini with mobile app and paid subscription.

  • The US Government creates the first-ever consortium dedicated to AI safety.


Business

OpenAI CEO Sam Altman seeks as much as $7 trillion for new AI chip project: WSJ (CNBC)

OpenAI CEO Sam Altman is reportedly in talks with several investors, including the government of the United Arab Emirates, for a project to increase global chip-building capacity. The supply-and-demand problems with AI chips are currently hindering AI corporations’ growth. The project would require between $5 - $7 trillion.

Google Releases Gemini, an A.I.-Driven Chatbot and Voice Assistant (NYTimes)

On Thursday, Google released Gemini, a smartphone app that replaces Bard and Google Assistant. Gemini is an all-encompassing digital assistant capable of nearly anything you could think of since it can respond to both images and sound. There is a free (but limited) version of Gemini but the more powerful Gemini Advanced is $19.99/month.

Over 4,000 workers have lost their jobs to AI since May, outplacement firm estimates—and that’s ‘certainly undercounting’ (Fortune)

According to an outplacement firm, 4,600 jobs have been cut since last May to either hire people with AI experience or replace the position altogether with AI. Companies like UPS, Duolingo, and IBM are just some of the names that have halted hiring or actively fired employees because of the integration of AI to improve productivity.

Government

Biden-Harris Administration Announces First-Ever Consortium Dedicated to AI Safety (NIST)

On Thursday, the U.S. AI Safety Institute Consortium (AISIC) was announced. The consortium is composed of more than 200 member companies and organizations that are at the forefront of creating and using advanced AI systems. Their goal is to support the development and deployment of responsible AI.

FCC makes it official: AI-generated voices in robocalls are illegal (Washington Post)

Following up on a story included in Monday’s newsletter, the Federal Communications Commission (FCC) has officially criminalized AI-generated voices in robocalls. Robocall scams using AI were already illegal, but the ruling on Thursday clarifies that the act of generating an AI voice for a robocall is illegal as well.

Science

New AI Circuitry That Mimics Human Brains Makes Models Smarter (Scientific American)

Mark Hersam, a chemist and engineer at Northwestern University, has been trying for years to mimic human brain computation in technology. Memory and processing are usually separate components of hardware unlike the human brain which combines both. Hersam and colleagues have recently redesigned transistors to function more like neurons. This will allow AI systems to move beyond simple pattern recognition. They will be able to form memories and make associations from them.

Entertainment

Holly Herndon’s Revolutionary AI Music (Jacobin)

Holly Herndon’s desire to sing with other people again prompted her and partner, Mat Dryhurst, to build an AI dataset based on both of their voices as well as a selection of close friends. The AI program, affectionately named Spawn, was treated more as a collaborator than generator for her new album. Herndon hopes to open the discussion that there might be a place for AI in music that doesn’t infringe on anyone’s rights.

Minecraft could be the key to creating adaptable AI (NewScientist)

Steven James and colleagues at the University of the Witwatersrand in South Africa have developed a benchmark test to measure the intelligence of AI models using Minecraft. The benchmark, titled Mineplanner, measures an AI model’s ability to think beyond it’s initial dataset to solve problems. They are hoping to eliminate the need for a human designer to give instructions for every task an AI might encounter. Rather, the AI will be able to infer solutions on its own.

Life

When A.I. Bridged a Language Gap, They Fell in Love (NYTimes)

When LeRoy Romero was set up on a blind date by a friend, it seemed like the perfect match except for one minor detail. He did not speak Spanish and his date, Brenda Ochoa, did not speak English. They used Google Translate for the initial date but later found even greater success using Captions, a video software that uses AI to create live subtitles or lip-dub. Lip-dub is a feature that can imitate a person’s lip movements and voice to make it appear as if they are speaking your desired language.


New at KPI

KPI Podcast | We’re recording our first episode this weekend! If anyone has any podcasting tips or would like to possibly be a guest, please contact us!


Chatbot Wisdom for Today

Want to learn something specific or ask advice? Submit your anonymous questions here and perhaps they’ll be featured in future newsletters!

Previous
Previous

Newsletter

Next
Next

Newsletter