Fortune reported on October 13 that some cybersecurity experts are skeptical of Anthropic's research on small models copying themselves onto everyday devices. The Register framed it as conjuring a threat that does not exist yet, one that depends on smaller models getting smarter.
The Register's coverage also noted Anthropic's framing is extrapolation, not a demonstrated worm. The disagreement shows how much of the AI risk debate turns on forecasting capabilities that do not yet exist, a point Dell and Huang made in their own ways this week.
Source: Fortune