AI Top-of-Mind 3.6.24 — To Your Health!
Top-of-mind is something running below the radar for most… an ongoing cyberattack against ‘Change Healthcare,’ part of UnitedHealth Group and responsible for billions of health-care transactions yearly. As ‘USA Today’ reports, it is impacting both providers and even individual doctors, and last year 1 in 3 Americans were impacted by health-care related breaches. More from HHS. Though there is no direct linkage to ‘AI-assist’ yet, it is only a matter of time.
Looking ahead, it gets scary, with researchers developing one of the first self-propagating worms that specifically targets AI applications. Never good when scientists develop something that should be contained. We know how that movie ends. From the ‘Infosecurity Magazine’ article:
· The new paper details the worm dubbed “Morris II,” which targets GenAI ecosystems through the use of adversarial self-replicating prompts, leading to GenAI systems delivering payloads to other agents.
· Once unleashed, the worm is stored in the retrieval augmented generation (RAG) and move “passively” to new targets, without the attackers needing to do anything further — something the authors described “0-click propagation.”
On the model front, more tips on running open source LLMs on the desktop, Gene Bernadin describes ‘Jan,’ available for download.
Thinking back to my reference yesterday of ‘I, Robot’ when describing ‘Figure.’ A good essay on how science fiction, for those who read it, has helped provide a foundation for lots of what is happening around us by imagining a set of possible futures. One good observation from Philip Harker’s writings:
By default, books, movies, and TV frame AI as the enemy. The AI is either not fit to self-govern or, more typically, is actively hostile against humanity. This is either expressed as revenge against their creators (as in the anti-robot jihad in Dune) or just the naturally destructive and colonialist nature of machine intelligence (ala the Borg in Star Trek). To say that most authors (with a few notable exceptions) have warned against AI for generations is accurate.
And turning to education, warnings by Tracy.3 writing in ‘A Teachers’s Life’ on overhype including the company ‘VIPKID’ as well as disturbing reports from China on student monitoring. The headbands have indicators depicting level of concentration.
And finally, most of us have smart TVs, and it was only a matter of time that the vendors would begin to market AI-based enhancements for image and audio quality including upscaling, but this is just the beginning. Using object identification, new models will be able to highlight items, and I imagine companies with paid placements causing their appliances to glow. More in the ‘tech radar’ article.
