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.
AI inside? Adobe introduces new credential system to identify AI content
Authenticity in the age of artificial intelligence represents a major challenge for content creators — including marketers — as chatbots produce increasingly realistic photographs, videos, voiceprints and other content. At a conference this week, Adobe introduced a potential solution: an embedded icon that identifies the content’s pedigree, including which AI software fueled its creation.
Digiday’s Marty Swant explains that Adobe’s new Content Credentials icon will embed in the metadata of content produced by Adobe software, as well as partner applications. Adam Rose, Chief Operating Officer of The Starling Lab, draws an analogy between Content Credentials and the “lock” icon signaling a website’s security. Though AI offers “infinite content,” Rose tells Swant, for marketers “there’s a need, and I would say a demand, for authenticated content. And that’s really what’s different here.”
Using generative AI for improved health outcomes
A new study highlighted in Healthcare IT News is among a spate of developments highlighting how artificial intelligence is poised to transform the patient experience in healthcare — a transformation that stands to rewrite the rules on how healthcare is delivered and marketed to consumers. Senior Editor Andrea Fox notes that the Generative AI Tracker, a collaboration between Pymnts and AI-ID, sees AI technology companies playing a major role in development of tools for improved diagnostics, treatment and delivery of care, while also enhancing the process of pharmaceutical development. She notes that the latest Generative AI Tracker report “covers use cases, offers examples of emerging regulations and highlights select startups making inroads to address complex challenges in healthcare settings.”
Meanwhile, Will Morton of radiology news website AuntMinnie.com considers a new study surrounding the potential uses of AI in the field of radiology. According to Morton, the study’s authors found that “evidence suggests that radiologists can use ChatGPT to ask questions about medical imaging, radiology procedures, and patient information, and receive accurate and relevant answers in real-time.”
Related news:
• “Google announces new generative AI search capabilities for doctors” (CNBC)
• “Microsoft Targets Healthcare with New Cloud-Based Data & AI Solutions” (Forbes)
Energy issues could pose sustainability challenges in AI adoption
Innovative technologies can bring unintended consequences, and for artificial intelligence, one of those may be the amount of energy it takes to run billions of queries. That’s according to Alana Mayham of Courthouse News, who writes about a commentary published recently in the scientific journal Joule.
According to Mayham, Alex de Vries, a Ph.D. candidate at Vrije Universiteit Amsterdam and founder of research company Dichotomist, notes that training large language models alone requires massive amounts of electricity — by de Vries’ estimate, anywhere from 324 to 1,287 megawatt-hours.
“For context, an AI tool consuming 433 megawatt-hours is enough to power 40 average American homes annually,” Mayham writes. In his commentary, de Vries says that Google’s search engine, with 9 billion searches a day, would consume 29.2 terawatt hours of electricity annually, representing a 60% increase from 2021.
Further reading
• BBC Will Block ChatGPT AI From Scraping Its Content” (Deadline)
• “Warner-backed generative AI startup from co-inventor of Siri launches own record label” (Music Business Worldwide)
• “Google’s AI Is Making Traffic Lights More Efficient and Less Annoying” (Wired)
• “US Space Force pauses use of AI tools like ChatGPT over data security risks” (Reuters)
Thanks for following along as we continue our coverage on the generative AI beat. See you next Friday.
Previously: “The Week in Generative AI: October 6, 2023 edition”
