A Powerful Tool with One Huge Missing Component
A lot of us covering AI have become increasingly concerned about the lack of control over what’s being developed and the haphazard and dangerous efforts surrounding it. For instance, this post about one of these experiments on Reddit had me wondering if we’ll make it to the end of the decade alive.
But a number of companies have stepped up to form the AI Alliance, a powerful group headed by IBM, Meta and a host of other businesses and educational institutions focused on creating better, cheaper ethical AI solutions that will assure a positive outcome for AI creation. But there’s one glaring omission: What happens to the displaced employees that will result from these efforts?
I was briefed this week on the AI Alliance, and I was impressed by its deep focus on ethics, helping companies and nations that are currently disadvantaged in terms of AI development to catch up, and assuring the result is efficient, high-quality and dependable.
However, right after that briefing, I participated in a podcast about job creation during this 5th (AI-driven) Industrial Revolution and realized there was a big problem that was being ignored. What can be done about the massive unemployment that will result since the new jobs we’ve been talking about are increasingly less likely to emerge?
Let’s talk about both things this week.
The AI Alliance: A Powerful Effort to Assure Ethical and Affordable Creation of AI
The AI Alliance is both new and impressive. Made up of most of the companies and educational organizations currently leading the AI charge, this alliance has six focus areas: the creation of AI skills and enabling academic research; building and assuring AI trust, safety, and in-depth evaluation; the creation of critical developer tools; diverse hardware enablement to build competition; the creation of open foundation models that will allow young companies and emerging countries to participate in the effort; and advocacy to assure the positive messages surrounding the technology will help its growth.
The AI Alliance has a powerful steering committee made up of some of the most capable partners. The committee focuses on progress, resource allocation and aggressively moves to find ways to unblock issues that prevent the ethical advancement of the technology.
IBM has long argued for the ethical use of AI, so its participation is critical to the positive outcome of this effort. IBM appears to be uniquely capable and focused on assuring a positive outcome while building collaboration, openness, and a rich competition-friendly environment in which to grow the technology.
I expect this effort will ensure that AI grows into a powerful tool for the good, which is a relief given the Reddit post I mentioned earlier.
However, what about the employees?
The Unaddressed Problem
When we spoke about the 5th Industrial Revolution being driven rapidly by AI development, we looked back to prior Industrial Revolutions that created new jobs that eventually more than absorbed most of the people willing to change who were displaced. However, this time, the tool is being used to fill roles that are understaffed. From Devin, an AI-created software developer, to McDonald’s, AIs are being focused on understaffed positions, and there is no position that has a greater chance of being understaffed than the coming need for AI experts who currently don’t exist but would typically take the retrained employees who were displaced by AI.
In other words, AI will be used to fill the jobs that AI creates because AI can more rapidly be trained up at scale than people can. Granted, it can take a few months to create a new model focused on a job category, but once that’s done, inference models that can then do the targeted job can be spun up in minutes or even seconds, while human retraining takes months to years. And most of the human retraining efforts haven’t even been staffed yet, let alone moved into production.
It is highly possible that by the time AI is massively displacing people, it will have already filled the jobs those people might have otherwise been retrained to take, resulting in equally massive numbers of unemployed and unemployable people.
Right now, AIs are already performing in-line with humans, having advanced nearly vertically over the last two years, and the tool’s advancement isn’t slowing down. By the end of the decade, humans may simply be uncompetitive, but large numbers of unemployed people can result in massive political and often violent upheavals. Machines only consume electricity, so if there are no people to buy products, what will the AI-driven industries do for income?
If we don’t aggressively address this potential unemployment problem, we are looking at a period of human unrest unlike anything we’ve ever seen.
Wrapping Up
The AI Alliance is a powerful consortium that will assure rapid advancement, ethical use, and a healthy competitive future for the technology. However, the continued lack of focus on what to do with the massive number of displaced employees that will result from advancements like Devin that will be used to fill the new jobs that are created initially and lock out increasingly non-competitive humans is even more problematic.
We need to envision a better AI future. Otherwise, the outcome will be similar to what this Twilight Zone episode forecasted, and no one’s job will be safe.
