Welcome to The Week in Generative AI, a weekly column for marketers from Quad Insights that quickly sums up need-to-know developments surrounding this rapidly evolving technology.
White House asks tech firms to manage AI
Executives from seven key tech companies in the AI space — Alphabet, Amazon, Anthropic, Inflection, Meta, Microsoft and OpenAI — met with President Biden at the White House on Friday. The goal of the meeting was to discuss guidelines and safeguards for generative AI platforms, including adding possible measures such as “virtual watermarks into the material they generate, offering a way to help distinguish real images and video from those created by computers,” according to Justin Sink and Anna Edgerton from Bloomberg.
There are still details to work out from all parties involved, as “the guidelines don’t prescribe approval from specific outside experts in order to release technologies” and that the “watermarking system still needs to be developed,” Sink and Edgerton report.
Related coverage:
- “The White House and big tech companies release commitments on managing AI” (NPR)
- “7 A.I. Companies Agree to Safeguards After Pressure From the White House” (The New York Times)
- “OpenAI, Google, others pledge to watermark AI content for safety, White House says” (Reuters)
Meta releases new AI tools
Meta, the parent company of Facebook, is in the process of releasing two AI tools, one for image creation/analysis and the other for language processing.
Their new text-to-image model called CM3leon (pronounced “chameleon”) can generate high-resolution images from written descriptions and write captions for existing images. This offering will compete with Adobe’s Firefly and Midjourney AI for visual AI rendering.
Jose Antonio Lanz at Decrypt writes that “In head-to-head comparisons, CM3leon appears to handle complex objects and constraints in text prompts better than models like OpenAI’s DALL-E 2, and even Midjourney.”
Meanwhile, in The Wall Street Journal, Tom Dotan and Deepa Seetharaman report that Meta is working with Microsoft to release Llama 2, a large language model, and that this “represents a shift for Meta, which just two months ago said it had no plans to make its model available commercially.”
Microsoft, one of the biggest investors in generative AI industry leader OpenAI, is clearly putting eggs in as many AI baskets as possible. “In making Llama 2 available to Azure customers, Microsoft is showing a willingness to broaden beyond ChatGPT maker OpenAI as its AI platform of choice,” write Dotan and Seetharaman.
Related coverage:
- “So, that’s the end of OpenAI’s ChatGPT moat” (Business Insider)
- “Meta Opens AI Chatbot Tech for Commercial Use via Microsoft” (Bloomberg)
- “Meta claims its new art-generating model is best-in-class” (TechCrunch)
- “Microsoft goes beyond OpenAI, makes Meta’s new A.I. model available to Azure customers” (CNBC)
- “OpenAI Worries About What Its Chatbot Will Say About People’s Faces” (The New York Times)
Apple is testing its own AI tool
Apple has started testing an AI framework code-named Ajax (though, let’s be serious, we’re all going to call it AppleGPT).
Bloomberg’s Mark Gurman offers some context: “Apple has been conspicuously absent from the [AI] frenzy. Its main artificial intelligence product, the Siri voice assistant, has stagnated in recent years.” As for Ajax, he reports that there is “no current plan to release to consumers” and that the company is “still trying to determine the consumer angle for generative AI.”
Related coverage:
- “Apple added $71 billion to its market value after news that it’s been secretly building an ‘Apple GPT’ to rival OpenAI” (Fortune)
- “Apple is testing an AI chatbot but has no idea what to do with it” (The Verge)
Real fake news? Bot-driven bylines are here
Google is testing a new AI tool, Genesis, that can write news articles, potentially upending the news industry, while also heightening concerns about the potential for AI to be used to create fake news.
Benjamin Mullin and Nico Grand at The New York Times, in reporting on Genesis, note that AI could enable users “to generate articles on a wider scale that, if not edited and checked carefully, could spread misinformation and affect how traditionally written stories are perceived.”
This follows recent coverage around G/O Media (owner of The Onion and Gizmodo) and the kerfuffle over error-ridden, AI-written articles that circumvented editorial control, as Sharon Knolle from The Wrap reported in June, as the news was “met with scorn by The Onion Union and GMG Union.”
Meanwhile, the Associated Press just joined forces with OpenAI to license archival news stories to ChatGPT. The AP’s Matt O’Brien says the news organization “doesn’t currently use any generative AI in its news stories, but has used other forms of AI for nearly a decade, including to automate corporate earnings reports and recap some sporting events. It also runs a program that helps local news organizations incorporate AI into their operations, and recently launched an AI-powered image archive search.”
Related coverage:
• “Google explores AI tools for journalists, in talks with publishers” (Reuters)
- “Gizmodo’s staff isn’t happy about G/O Media’s AI-generated content” (The Verge)
- “Thousands of authors urge AI companies to stop using work without permission” (NPR)
Further reading
- “School’s out for summer, and if ChatGPT usage is falling because students are off, that’s a bad sign for OpenAI” (Yahoo Finance)
- “Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds” (Fortune)
- “OpenAI vice president denies they’re making ChatGPT stupider” (Futurism)
- “A.I. is a $1 trillion investment opportunity but will be ‘biggest bubble of all time,’ CEO predicts” (CNBC)
- “Microsoft Teams is adding Maybelline’s AI-powered ‘makeup’ filters” (The Verge)
Thanks for following as we cover the generative AI beat. See you next week.
Previously: “The Week in Generative AI: July 14, 2023 edition”