With MAICON 2023 just around the corner, Paul Roetzer and Mike Kaput break down very different directions of AI this week. From incredible to dumb, from thorough to questionable, there’s lots to break down.
Meta’s incredible new (free!) ChatGPT competitor is here
Meta’s latest announcement has big implications for the world of AI. The company announced that its new, powerful large language model, LLaMA 2, will be available free of charge for research and commercial use. The model is “open source,” which means anyone can copy it, build on top of it, remix it, and use it however they see fit. This puts an extremely powerful large language model into anyone’s hands—and gives them the appropriate permissions to build products with it.
But that’s not all. It signals a major strategic direction that Meta is taking to compete with other AI companies—one that could have an effect on AI safety. Some major AI players place serious restrictions on the use and release of their models, often due to concerns about how models might be misused if they’re put in the wrong hands without guardrails. Meta is taking the opposite approach, believing that getting the technology into anyone and everyone’s hands will make the technology better much faster—and more quickly help Meta reveal and address issues that contribute to safety, like the use of the model to produce misinformation or toxic content. Will this new approach be successful? 
Elon Musk changes Twitter to X
Musk is in this news again. As of the morning of the podcast recording (July 24, 2023), he has formally rebranded Twitter as X. The platform formerly known as Twitter hasn’t changed much aside from its logo, but it seems like Musk and leadership are viewing it as just one piece of a much larger entity. 
In a somewhat cryptic set of tweets, CEO Linda Yaccarino said: “X is the future state of unlimited interactivity – centered in audio, video, messaging, payments/banking – creating a global marketplace for ideas, goods, services, and opportunities. Powered by AI, X will connect us all in ways we’re just beginning to imagine.” In the past couple weeks, Musk has also announced xAI, his new company dedicated to building “good” artificial general intelligence and competing with OpenAI, among others. Time will tell what this means for the future of the brand.
Meta, Google, and OpenAI Make AI Responsibility Promises to the White House
Seven major AI companies—Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI—have all agreed to safety commitments proposed by the White House.
The commitments include promises to engage in “security testing” carried out by independent experts, using digital watermarking to identify AI-generated vs. human-generated content, testing for bias and discrimination in AI systems, and several other safety-related actions.
It should be noted these are simply voluntary commitments publicly announced by the companies and the White House, not any type of formal regulation or legislation. We discuss this on this week’s episode and will keep our eyes on this for you.
Tune into the last pre-MAICON 2023 episode! We’ll be back next week with more news, more insights, and lots of MAICON takeaways to share with you all.