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.

Meta to launch new AI chatbots and text-to-music AI tools

Meta, the parent company of Facebook, is planning to launch a range of artificial intelligence chatbots that can take on different personalities. According to the Financial Times, which broke the news, these chatbots, known internally as “personas,” might launch this fall and could be used to provide search functions and serve as a recommendation engine.

These chatbots would be able to collect vast new amounts of data on users’ interests, which could help Meta better target users with more relevant content and ads. The FT quoted Meta founder Mark Zuckerberg from an earnings call in which he said a new generation of AI chatbots could serve as “assistants” or “coaches” and could “help you interact with businesses and creators.”

Meanwhile, on Wednesday, Meta announced it is open-sourcing a text-to-music suite of AI tools called AudioCraft, which will allow content creators to “input simple text descriptions to generate complex audio landscapes, compose melodies, or even simulate entire virtual orchestras,” as Benj Edwards reports in Ars Technica.

Related coverage:

“Meta is reportedly preparing to release AI-powered chatbots with different personas” (TechCrunch)

“Meta’s Ray-Ban Smart Glasses Fail to Catch On” (The Wall Street Journal)

“Instagram seems to be working on labels for posts ‘generated by Meta AI’” (Engadget)

AI hacking draws fresh concern

A growing concern surrounding artificial intelligence is the rise of AI-generated cybercrime and malware, as Jason Nelson reports in Decrypt. Citing the research of cybersecurity firm SlashNext, Nelson writes that cybercriminals are deploying AI tools that have “the ability to create phishing scam web pages, write malicious code, create hacking tools, and write scam letters.” WormGPT and FraudGPT, two such malicious chatbots, were discovered by SlashNext in their investigation, with more on the horizon.

Meanwhile, writing for Wired, Will Knight shares research from Carnegie Mellon University showing that “adding a simple incantation to a prompt — a string text that might look like gobbledygook to you or me but which carries subtle significance to an AI model trained on huge quantities of web data — can defy … defenses in several popular chatbots at once.” The concern with chatbot manipulation is that while major AI companies pledge their allegiance to guardrails on their technology, “they have not figured out how to block adversarial attacks more generally.”

Related coverage:

“The generative A.I. battle between companies and hackers is starting” (CNBC)

“FBI Warns About China Theft of US AI Technology” (Voice of America)

“‘Wonder and Worry’: How Biden Views Artificial Intelligence” (The Wall Street Journal)

Google Assistant to get an AI enhancement

Google is in the process of reshaping its Assistant product — starting with the mobile version — to incorporate generative AI, according to a scoop from Axios’ Ina Fried. Summarizing an internal Google email, Fried writes that “Google is reorganizing the teams that work on Assistant” and is “making a small number of layoffs in conjunction with the shift.”

Over at Ars Technica, Ron Amadeo reads the tea leaves on the development, concluding that this is the next phase in the voice-assistant battle between the biggest players in tech. “Google and Amazon’s Alexa started this voice assistant war,” he writes, “and are now in the same boat: products that were loss leaders years ago never discovered a revenue path and have run out of runway. … Both products were created when voice assistants were the hot new thing, and now that AI language models are the hot new thing, pivoting buys them more time as loss leaders to hope to someday find a revenue stream.”

Related coverage:

“Google will ‘supercharge’ Assistant with AI that’s more like ChatGPT and Bard” (The Verge)

“Leaked email shows Amazon formed a new group to work on the ‘most ambitious’ AI models” (Business Insider)

“How SEO Agencies Are Adapting to a World Without Traditional Search” (Adweek)

Further reading

“Chinese tech giant Alibaba challenges Meta with open-sourced A.I. model launch” (CNBC)

“AI-Ready Data Centers Are Poised for Fast Growth” (The Wall Street Journal)

“Dell is all in on generative AI” (The Verge)

“How generative AI has shown up in earnings chatter again this quarter” (Digiday)

“Tim Cook touts that Apple has been investing in generative AI tech ‘for years’” (9to5Mac)

“The AI rules that US policymakers are considering, explained” (Vox)

“Microsoft-backed OpenAI files trademark for ChatGPT powered by GPT-5” (Windows Latest)

“‘It’s destroyed me completely’: Kenyan moderators decry toll of training of AI models” (The Guardian)

Thanks for reading along as we cover the generative AI beat. We’ll see you next week.

Previously: “The Week in Generative AI: July 28, 2023 edition”