Recent concerns about the safety of AI may be overblown, according to multiple Pitt professors.\u00a0<\/p>\n
In recent weeks, a former Anthropic employee expressed concerns<\/a> in his resignation posted to X that AI development could get out of control and lead to disastrous results. The incident has sparked conversations<\/a> about the potential dangers and regulation of AI.\u00a0<\/p>\n Patrick Skeba, a teaching assistant professor in the Department of Computer Science, said fears of AI being an \u201cexistential threat to humanity\u201d are \u201ccomplete nonsense.\u201d<\/p>\n \u201cSaying that AI is dangerous implies that it can somehow do things that we don\u2019t ask it to,\u201d Skeba said. \u201cWhen they say things like, \u2018What if it brought us to war? What if it did all of these things?\u2019 Well, why would we give it that capability?\u201d\u00a0<\/p>\n Adam Zweber, a teaching assistant professor in philosophy, said fears of AI destroying the world are not new, citing the existing \u201cprobability of doom<\/a>\u201d metric, colloquially known as P(doom).<\/p>\n \u201cThere are lots of ways in which things can go very bad, even if AI isn\u2019t that much smarter than us,\u201d Zweber said. \u201cThere are lots of risks that can lead to very bad outcomes.\u201d<\/p>\n Zweber explained that people within companies like Anthropic<\/a> or OpenAI<\/a> work to prevent the possibility of catastrophic events.<\/p>\n \u201cIn principle, you could build an AI without guardrails,\u201d Zweber said. \u201cBut there are a lot of people working on this \u2014 a lot of very smart and very concerned people who are coming up with ways to prevent it from just telling people how to do bad things.\u201d<\/p>\n Skeba said situations like the Hugging Face incident<\/a> don\u2019t point to a disaster being caused by AI. In these situations, AI agents are given internet access and break out of \u201cpoorly configured sandboxes,\u201d according to Skeba.<\/p>\n \u201cYou basically gave them the key and said, \u2018Promise not to use this,\u2019 but they used it anyway,\u201d Skeba said.<\/p>\n Skeba said there are reasons to be concerned about the potential usages of AI, but these worries lie in human, rather than artificial, decision making.<\/p>\n \u201c[AI] can have some really unjust and terrible results,\u201d Skeba said. \u201cBut to blame that on the AI doing that of its own accord, as opposed to saying, \u2018Why are we taking human judgment out of this loop and replacing it with AI?\u2019 That\u2019s very different from the AI somehow choosing or deciding to do things that humans don\u2019t like. That\u2019s science fiction in my opinion.\u201d<\/p>\n