The Software Revolution: Societal Impacts, AI Risks, and the Rise of Small Organizations

The three technological revolutions in human history: the agricultural revolution, the industrial revolution, and the ongoing software revolution. Sam Altman believes the software revolution will cause massive unemployment and empower small organizations.

Great technological revolutions have affected many people and social structures. The previous industrial revolution created many jobs because new technologies required many humans to support them, but this time is unusual; it defies the conventional belief that technological revolutions should always benefit the economic status of the majority.

Like other revolutions, the software revolution creates wealth but also causes job loss. While new things will satisfy humanity's infinite needs, we need not pretend about the software revolution: it will benefit the middle class (the rich get richer).

Technology provides a balance of capability and luck, condensing wealth and generating inequality. Extreme wealth inequality will be the biggest social problem of the next 20 years. We can redistribute wealth, but that does not solve the underlying issues people truly need to address.

Trying to cling to meaningless work is frightening, yet it is a popular idea. Trying to find new jobs for billions is a good idea but difficult, regardless of what those new jobs are. They will be fundamentally different from anything existing today, making meaningful planning impossible. Current strategies are wrong; when Travis talks about autonomous cars and Uber (like Didi) creating millions of jobs, he is joking.

The second challenge of the software revolution is the concentration of power in small organizations and companies. Although this also occurs in most technological revolutions, recent terrifying technologies like the atomic bomb taught us a bad lesson. Meanwhile, the industrial revolution increased jobs; building an atomic bomb is difficult because technology is limited—it requires massive energy and enriched uranium, demanding national resources.

This time is different; the software revolution proceeds in the usual way of technological revolutions but makes small groups stronger.

The two biggest risks of the software revolution are AI and synthetic biology. Placing huge capabilities in the hands of small organizations could cause harm; in small labs, terrifying diseases could be designed. AI development could end human life, as perhaps only a few hundred people worldwide would need to work in offices, needing only laptops.

New survival challenges will not require national resources. Previously only a nation could build rockets; now a company can accomplish at least part of it (largely through software), yet a rocket can destroy anything on Earth.

What can we do? We cannot declare the knowledge to make these things illegal; instead we should hope it continues to develop. We cannot stop technological progress.

The best strategy is to specify wise safety measures, striving to ensure the benefits of technology outweigh its harms. If we can synthesize new diseases, perhaps we can also synthesize vaccines; if we can create bad AI, perhaps we can create better AI to counter it. (The higher the magic, the higher the virtue.)

But the current handling strategy is seriously wrong. Although not like dealing with atomic bombs, the sooner we stop ignoring and pretending (like an ostrich burying its head), the better our future will be. In fact, we have not had any serious debate on synthetic biology and AI, which is surprising.

Clearly, I love the software revolution, I am lucky to be born in this era, but I worry we have learned the wrong lessons from recent cases. The two situations described earlier—massive job loss and concentration of power in small organizations—should not escape our attention.

(Source: Concurrent Programming Network)