AI may or may not be the most important technology ever invented, but it is by far the most rapidly deployed.
Technologies like the mechanical clock weren’t widely adopted for at least a century. It took almost a decade after the iPhone’s introduction for half of American adults to slip a smartphone into their pocket. AI is moving so fast that the time between theoretical breakthroughs and commercial availability is measured in weeks.
There seems to be no way to effectively place ethical guidelines on tech moving at this velocity, and the AI Action Summit that took place in February in Paris made it clear that we’re not even going to try. Because countries see AI as a part of their national security, they are loath to voluntarily or unilaterally slow down its development.
This has left us mortals in an impossible situation. Even if you do not believe that artificial general intelligence (AGI) and artificial superintelligence (ASI) are existential risks to humanity, the widespread availability of such powerful tools has given individuals a new way to hurt others from a distance.
AI has the potential to massively improve human welfare, but its proponents are often short on details about how exactly this will happen. In the meantime, the potential for abuse is increasing daily and there are few mechanisms in place to prevent bad actors from using AI to harm the general public or specific individuals.
What in the world are Jews — both leaders and laity — supposed to do in this situation?
I firmly believe we must do something. Judaism’s commitment to cultivating an ethical society demands no less. But time is limited, and Jews with busy schedules will rightly look at this topic and wonder whether addressing it — which involves keeping pace with its many twists and turns — is really the best use of our time. If political and business leaders struggle to wrangle this technology, why should we expect a minority religion to contribute anything?
Let’s be clear: We should not expect Jewish thought to sway AI development directly. This global project will hardly face regulation through secular argument, let alone passages of the Talmud.
While Silicon Valley’s tech culture may be particularly secular — or perhaps its religion is tech itself — Jewish thought is nowhere near sufficiently equipped to provide strong ethical guidance, and its ability to shape American secular society is limited at best.
Instead, Judaism’s greatest power in this technological maelstrom is our community. This is not because of some romantic notion that we are stronger together. It is because you can’t regulate technology effectively while ignoring what communities actually want.
This simple truth seems to have evaded most tech regulators even when regulation was politically fashionable. For many tech developers, the public exists only as a willing receptacle for new products, an abstract concept called “scalability.” This is sometimes because tech produces real good, but a lot of the time it is because the public has grown used to the idea that 21st-century citizens must adopt tech regardless of its utility. Tech comes in waves. When a new one comes, you either surf it or you drown. It happens to you, not for you.
This learned sense of inevitability means that the public has lost its ability to evaluate new tech in terms of its own vision for society; the muscle is so atrophied that average citizens don’t even realize this is something they can do. Instead, this task ends up getting outsourced to policy experts. Yet policy experts can only implement societal values when those values have actually been articulated.
AI ethics, even on a theoretical level, is patchy. Experts can develop policies for values that are clear — don’t encourage self-harm, don’t enable exploitation, don’t violate privacy — but they’re not equipped to resolve our general anxieties (about losing your job, about romantic relationships with robots, about an internet full of auto-generated nonsense). AI ethics, therefore, ends up being reactive and is better at solving small technical problems than big societal ones. It also means that the public’s voice gets mediated by a third party.
This is where I think religion in general, and Judaism in particular, can do some good.
To the extent that they are champions of ethics, religious communities are more than just places where ethical philosophy gets a sprinkling of Bible verses. Religious communities are places where ethics itself is inculcated, where both children and adults can learn together what a just society ought to look like. This process is about more than just the application of age-old laws to novel situations; it’s about the messy but necessary process of making determinations about the world we wish to inhabit.
If this seems difficult, consider that the alternative is for someone else to make that determination, which means resigning ourselves to a reactive ethics that only speaks up when things are so egregiously (and probably irreversibly) wrong that we feel we must at least voice our dismay, even if doing so is futile.
Much is made of the need for AI to align with our values. Religion is an important component of human alignment: a method to first determine what actually matters and then to ensure that the fastest-moving tech does not stray too far from what we understand about our newly articulated values.