By Thiyagarajan Maruthavanan
Post gun adoption, those still fighting with swords and sticks quickly became enslaved and colonized. The recent platform shift arising in AI is one such technology inflection point. Every country is deliberating its choices.
Don’t fake sloganeer, get your country comparative fit
Fund allocation and a new central-level policy are typical governmental responses to any change. States emulate by creating economic zones. This is bow tied with a slogan. This is the equivalent of a fake CEO stuttering that their company is AI-first. Such sloganeering doesn’t even win political campaigns. More deliberation is required. Think beyond one election; extend your horizon to 10-20 years. History remembers those who handheld humanity through key inflection points.
Post GPT era is that inflection point.
One question on every leader’s mind today is.
Should you build your LLM?
The last decade demonstrated that platform wars are also geopolitical, permeating national boundaries. Twitter has shown it can help overthrow elections. Had nations known earlier, they would have slapped sanctions in the initial years of its existence, but now it’s too late. Nations fear a platform becoming as powerful as a nation-state. The successful ones dominate multiple geographies. For example, investors tried to fund Google and Facebook in India. In a few years, it became evident that Google in India was Google. Facebook in India was Facebook. China, of course, was an exception where the state created Baidu as an alternative.
Therefore, nations today ask if they should build their platform to retain power during this shift.
Wrong question. The right one is what Sun Tzu teaches us to ask:
To deal with any conflict, know yourself and your conflicting neighbor, focus on your strengths, and attack their weaknesses. We are in the age of instant research sharing, where orxiv.org ranks how many AI tinkerers published papers from your country. That should give you an idea of your LLM strength. Every country desires abundant crude, like Middle Eastern countries. However, some countries gain an edge through natural selection or past path dependence. Therefore, the question is, with the re-organized value chain by LLMs, what’s the most advantageous place for your nation? For instance, for India, it would be AI services, not AI platform infrastructure.
Inspire the unique opportunity focus area
According to OpenAI’s report, 19% of jobs will be 50% impacted, and 80% of jobs will be impacted by at least 10%. Every nation must relook at mass population reskilling. Changes like these trigger painful migration that profoundly impacts society. Furthermore, many political wins can be scored in this battle. Contemplate how reskilling can become a key action in the election manifesto.
Investing in education from primary school to university will be necessary but insufficient. Rather than threat awareness, a mass’s competitive spirit often drives them to action. Inspire a Sputnik moment that launched the space race for the US. For India, it might be Sam Altman’s comment that Indian startups may be unable to gain an AI edge on small investments. But to reiterate, choose an area where you have a comparative advantage.
Find a moment to stir your citizens into motion while providing resources, concessions.
Technology always rescues capitalism. It saves capitalism from the capitalist. History shows that wealth distribution is dictated by power law. Technology facilitates the change to sit on the top of the power law. The top 5 companies in global stock indexes today are tech giants.
Ignore the false smoke, focus on real society level harms
Safety of humanity is a keyword that will grab media attention and aid political sloganeering.
Our mass understanding of AI safety stems from movies like Terminator & Eagle Eye. That depicts AGI. Point of singularity where machines become capable of enslaving humans. Or that power will allow human dictators to oppress marginalized sections. Every century has doomsayers and alarmists. If recent COVID tackling is any indication, society always solves humanity wiping level threat problems.
These doomsayers take away attention from real problems that can affect nations adversely- Regulatory capture- when businesses develop innovation; they create a moat around it to enjoy profits for extended periods. The most powerful and oldest type of moat is regulatory. Businesses work closely with the government to get favorable laws enacted for them. Cory Doctrow has elucidated how tech companies have aced regulatory capture and fine tuned their playbook into three different types:
- Face in power- where strongly shaped public opinion gets favorable regulation.
- Revolving door- where industry execs become a part of the policy-making process, best illustrated by how Net Neutrality was killed in the US.
- Shape of the debate- defining axioms. Think of the self-fulfilling idea that government projects are “wasteful” and “inefficient.” Once all players accept this idea, the debate shifts from “What should the country do?” to “Which private entity should the government pay to get it done?”
The first is like winning the game, the second is like refereeing, and the third is like controlling the game’s existence and rule book.
As AI changes are poorly understood, the leading tech companies are playing level 3. Policy works’ are meant to protect citizens and not offer moats to businessmen that hobnob with politicians.
Steps to not get shuffled out of the power law
Don’t ban or create licensing. Technology change cannot be curbed by rationing. Banning something only makes it grow faster. Regardless of your cultural roots of distributing change, command and control or laissez-faire, attracting top talent has significant advantages. It can even help tackle structural disadvantages. Is your immigration policy rolling out the red carpet for AI talent? At least don’t repel them.
Inspire the nation in your AI focus area by tapping into the competitive spirit. Appoint academics rather than industry players in regulatory roles.
Power struggles are won in the mind before they are fought. Making a deliberate choice on what is changing and what gives you strength in the new world of AI will ensure that you don’t get shuffled out in the power law shaped by AI.
(The author is Thiyagarajan Maruthavanan, Partner, Upekkha, and the views expressed in this article are his own)
