Written by: Zeev Farbman
The panic of AI killing humanity is hysteria and is distracting us from the very real risks that do exist, largely human-driven. Reason, thoughtfulness and rationality are the only approaches that will take us forward.
The idea that we will put the brakes on technological development is foolish, not because it is right or wrong, but because it will not happen. So we need a smarter plan. One that focuses on how the technology gets misused by humans, not whether it autonomously turns against us.
Rather than lament the rise of AI, we should be full speed ahead on identifying vulnerabilities bad actors could exploit using the technology, and address them quickly.
Just as powerful technologies of the past, like nuclear, carried the risk of catastrophic events, AI has very real risks. They are not, however, that AI will become conscious, or even achieve full autonomy without sentience, and turn against us. The risk is that the technology will become incredibly powerful and humans will use it toward malicious ends.
We should welcome all points of view to the table, but the brilliant engineers working on this technology do not necessarily have a monopoly on perspective or truth. One tweet should not send society into a full blown panic.
We already know everyone will not suddenly slow the development of AI. If all of the American companies stopped developing AI tomorrow, it would race ahead in China or Russia or the Middle East. This is not like nuclear technology, where the ability to develop it was so complicated. Nuclear research relied on extremely rare natural resources, as well as hard to access technology, making it easier for a small number of governments to limit its spread.
AI development rests partially on access to leading GPUs, but even the most advanced types are massively more available than elements needed in nuclear production, like enriched uranium or plutonium.
Simply put, the notion that we can slow the development of AI is false.
Few even seem to believe it themselves. The same people claiming to be slowing the development of AI are the same ones rushing ahead with new models and trillion dollar IPOs. If executives are not giving up that capitalization of the technology, then do not believe that they are serious about curbing its development.
There is definitely positive work currently being done on misuse and mitigation of potential failures within our critical infrastructure. That work deserves support.
This is very different from saying we should, or could, slow down development of the underlying technology, which is a distraction from this more useful work. We cannot realistically press a button and slow it down. However, we can control how exposed we are when people try to misuse it.
The efforts we should be doubling and tripling down on are the real risks: how humans can use AI to exploit gaps in our technology in areas like banking, hospitals, supply chains, and other critical areas of society.
A sustainable path forward is doing the hard work of systematically identifying and addressing problems in our critical infrastructure. It’s not a great hot take for the social media cycle, but it is exactly the approach that has helped secure the world from previous technology waves, like nuclear, that presented enormous risks.
Related: You’re the Family CFO: How to Manage Money Across Three Generations
