Will the AI frenzy add up to anything?

Decades ago, a prominent technologist approached us with concerns about the potential dangers of artificial intelligence to human survival. However, at the time, the concerns of the Trump and Biden administrations were taken seriously. Lately, major labs have been warning against releasing their research models due to the danger they pose.

The government has been reviewing proposed releases since June, and two cases - Anthropic's Fable 5 and Mythos 5 - have already been addressed with heavy-handed measures. The same is true for Dario Amodei of Anthropic, who called for a cautious approach to the field, and Sam Altman and Elon Musk, who agreed to pace the frontier because they are already aware of the risks.

The frenzy surrounding AI is passing, but the debate over AI consciousness and rights is still ongoing. The U.S. and Chinese governments are working to maintain their authority despite AI. The transition from chatbots to active agents will be gradual, and the roadblocks to superintelligence include power supply, chip supply, and the lack of human-like qualities like tacit knowledge and muscle memory.

The AI analyst Tom Reed reminds us that even a 100% success rate at predicting the tokens of a Robert Caro biography doesn't generalize to becoming POTUS. Power supply and chip supply will continue to be challenges, and AI is out of control globally. We must continue to develop AI to understand its capabilities, but it remains to be seen if today's U.S. research can be commercialized or if investors will be swayed by Chinese freeware. The national security state may eventually take over top-tier AI.