Be warned about generative AI — a modern-day Trojan horse
In the name of improving our lives, enthusiasts posit the onset of AI as the solve-all for the world’s most pressing problems. This group includes Bill Gates. In his Gates Notes newsletter dated March 23rd, appropriately titled “The Age of AI has begun” for all the love he shows this nascent new tech, he says that he’s “…been thinking a lot about how — in addition to helping people be more productive — AI can reduce some of the world’s worst inequities.”
I’m sure the PC-pioneer and tech billionaire-turned philanthropist means well, right? Never mind the fact that he himself is heavily invested in the success — and profitability — of OpenAI, the developer behind the artificial intelligence that powers ChatGPT.
Gates goes on to list a few industries and areas that could directly benefit from AI: health care would become “better” if patients could be screened by a computer to determine whether they need medical attention from a human. The fight against climate change could be helped by leveraging crop data analysis in poor countries, and education could be “improved” by reducing face time between students and teachers in favor of a tailored learning platform, even though Gates admits that computers haven’t had the profound effect on education that people in his industry hoped for.
Gates largely glosses over the risks associated with AI. However, he says that “AIs will need to be trained on diverse data sets so they are unbiased and reflect the different cultures where they’ll be used”. In my mind, the data set is not the issue. The training data comes from the internet, a data set that is very diverse already. Admittedly, it is not diverse enough, as it only covers about half the planet, but it still covers a wide range of peoples and cultures. The issue is humanity, and that should be fairly obvious just from the brief few months that anyone and everyone has had access to generative AI. The algorithms inflate and explicate the worst in us.
Let’s get one thing straight: technology is never inevitable. This is just the narrative we’re being fed. Much like Sigal Samuel writes in her article “The case for slowing down AI” in Vox, there are lots of technological innovations that we decided not to pursue further, or to strictly regulate in order to manage them properly: “…the kind of innovations where we need to balance substantial potential benefits and economic value with very real risk”.
Yesterday, on March 29th, over 1,000 experts in the field of technology published an open letter, calling for a global pause on the rollout of “giant” AIs for (at least) six months, so that the risks can be properly defined and mitigated. The letter notes that this “does not mean a pause on AI development in general, merely a stepping back from the dangerous race to ever-larger unpredictable black-box models with emergent capabilities”. This is vital, as I believe AI can, if applied correctly, be valuable. But beta testing increasingly powerful generative AIs on the general public, each one a more intricate version of a modern-day Trojan horse, is not the way to introduce this technology.
Our economic system is set up to seek more output from each person, and AI, like the technological advancements that came before it, is both being touted as a way to release humans from mundane tasks and as a way to increase productivity. As always, the productivity aspect usually wins — it already has the upper hand with generative AI. And while this technology will change our lives, the implied improvement is still very much up for debate. With the first wave of AI-chatbots destroying fledgling careers of creatives and flooding the internet with auto-generated trash, perhaps eroding lives would be a more fitting description?
