Global Hype Cycle About AI is On
The role of government in the roll-out of artificial intelligence
A couple of weeks ago, Bernie Sanders, an outlier in the Senate but a man usually on top of things that concern workers, introduced a bill to cut the work week to 32 hours. The bill won’t go anywhere because the Republicans control the House of Representatives, but it acknowledges that in many fields, a 40-hour week is no longer necessary. Is Bernie just a voice trying to prepare America for a dramatically altered work environment as AI rolls out? What jobs will AI actually take, and when?
The hype about AI has been global, and governments have become aware that they might actually have to prepare for it, given that the task of any government is to keep its citizens safe. But what does safe mean in a radically new environment that hasn’t actually appeared yet? And what does that mean in the case of industrial policy?
Reminder: When governments get ahead of a crisis, they can make tragic regulatory or legislative mistakes. Fortunately that seldom happens. In the US, Open AI’s visionary leader Sam Altman has made several trips to Congress to familiarize that august body with the way AI could work. But Congress, with more pressing things on its mind, filed Altman’s information on a top shelf somewhere.
A couple of other governments, such as Ukraine’s, Russia’s and Israel’s are still preoccupied with last year’s existential issues, namely wars. They’re thinking more of artillery than of AI.
But some governments with a bit more leisure to contemplate are actively deploying various initiatives to prepare their citizens for the disruptions they suspect will be brought about by AI, focusing on re-skilling, education, and fostering innovation. These initiatives range from subsidized training programs to international collaborations aimed at harnessing AI's potential while ensuring a fair transition for the workforce.
Finland, through Helsinki University, has created a course in AI for its citizens and put it on the internet for everyone and Singapore's SkillsFuture program is a “comprehensive initiative offering citizens up to SG$500 annually for approved retraining courses,” with particular emphasis on technology-related fields. This program has already seen significant participation and is especially beneficial for individuals over the age of 40, offering subsidies of up to 90% on training costs. In Singapore, if you want to retrain and can figure out what to retrain in, or what to retrain for, money is not your deterrent.
In the UK, a National Retraining Scheme was launched with an initial funding of £100 million to address future work changes, focusing on adults 24 years and older without a degree and in low or medium-wage jobs. The scheme includes online and in-person training to improve technical skills, supported by national career service advisors. But how are those career service advisors going to know what jobs will last?
France's Compte Personnel de Formation (CPF) allows employees over the age of 16 to accumulate hours for training courses, ensuring paid leave during their training. This initiative has seen a significant increase in participation, with a wide range of courses available, including language and IT certificates.
Americans are talking.An American Workforce Policy Advisory Board was established to create education, training, and skill-building opportunities, involving leaders from various sectors to address workforce issues and promote lifelong learning.
Moreover, the OECD highlights efforts by countries like Canada, Germany, and Egypt working to empower citizens and equip them with necessary skills for AI-enhanced workplaces. These national experiences involve strengthening formal education programs in AI and STEM, promoting lifelong learning, and developing policies to attract global AI talent.
These programs all look to be saying “you may have tried to avoid technology now, but it’s coming for you.”
Only Germany appears to have a strategy beyond just making people go back to school. Germany's AI strategy includes launching transfer initiatives, digital test beds, and regulatory sandboxes to speed up AI innovations, with focused research and support in healthcare, environment, climate, aerospace, and mobility. Germany also emphasizes international cooperation, aiming to align AI development with the Sustainable Development Goals and foster global networking and collaboration in AI.
With the exception of Germany’s plans, all the others seem to be little more than glorified skills development programs. But what skills are necessary for a world enhanced by AI? We don’t yet know. The large language models we are testing now reveal more about the limits of AI than its successes. In certain fields, like drug development and government itself, the LLMs can help streamline some tasks. But take your job? Not yet.
So far my personal experiments with AI have been largely fruitless. It can’t plan a trip for me, or tell me what groceries I need, or fill out my income taxes. What can it do? Research (although you have to check it.) Create mediocre visuals, templated songs, or first drafts. A firehose of mediocrity.
AI is developing very quickly, its creators assure their investors. Next year there will be fewer “hallucinations.” But those movies where the robots take over the world? Not quite yet. Creativity is not math. It is not an equation. It is a gift, and we don’t even know which part of the human brain it comes from yet. So how can we re-create it with “compute.” ? First we need more neuroscience research.
We can’t reproduce creativity yet.. But the convergence of neuroscience studies with AI studies suggests we may get there some day. And then we will need a much smarter government than we have now to help us.
I know people that have been replaced by AI. Typically, they have been those that are extraordinarily well compensated. Analysts at Goldman, and top execs at IBM.