The guns of August are set to fire in the AI battlefield

date

Feb 8, 2023

slug

au-battlefield-august

status

Published

Meta

ChatGPT

type

Post

ogImage

summary

OpenAI started a new Gold Race between the candidates to rule the future at the expense of the billions.

After a few years without complete clarity over the next war scenario in tech, artificial intelligence won all the bets, with all major tech players forced to accept that it would become the kingmaker and the kingslayer of the industry for the next development cycle. The question is: how will the godfathers of technology be affected by the next war?

ChatGPT, the latest industry trendsetter, has captured the hearts and minds of users worldwide. It may not be perfect, but to use an expression coined by late MIT professor Clayton Christensen, it is good enough to do the job of the average Joe. All products to be released, such as Google's Bert (Bidirectional Encoder Representations from Transformers), will need to accept the terms of engagement imposed by OpenAI, and even more so when ChatGPT 4 is unleashed.

Google knows they screwed up. Bert’s timing for launch was wrong, and it seems they will drop it in the jungle before they had planned to. They should be worried. OpenAI's disruptive AI crusader is the first serious threat to Google's money-making machines, the search-driven apps under the Google Search umbrella. General William Garrison, the prime commander of US forces in Mogadishu portrayed in "Black Hawk Down," says, "We have just lost the initiative." That's exactly Google's case.

But Google didn't drop the ball alone. Mark Zuckerberg did it too. Meta's CEO had already admitted the then-Facebook error regarding video in social media two years ago. Now, after a couple of years burning money into the metaverse and making his shareholders uneasy, he finally admitted that AI is the new black, and it will have the whole focus of Meta too. Meta, though, explores a different spin on search and is less vulnerable than Google, thanks to the immense walled gardens of Facebook, Instagram, and WhatsApp. The question is: will the metaverse project become the Google Plus Zuckerberg's failure, or can he buy his way back in with a major acquisition? OpenAI is hardly for sale when it's riding a tsunami-like wave, but its theoretical US$30 billion market valuation shows that this is still a toy the big kids can think about buying.

And there is Twitter. Once Elon Musk is no longer experiencing the Goebbels-like amount of attention he had after capturing Twitter, his new plan is to make Twitter become a giant troll similar to Chinese WeChat, where users barely leave the app so stuffed with features it has inside. Such a venture has the potential to give him the exposure he craves, but in the short-term, it's a risky bet. Amazon, IBM, and Apple will fight with AI a bit away from ChatGPT's hit-zone. Microsoft, the least trendy of the tech megacorps, lies in silence because it has a relevant stake in OpenAI, so it's hard to imagine Satya Nadella hushing anything out of the oven.

One second last afterthought: what about China? Chinese capabilities in the AI world are not behind at all. About four years ago, Amy Webb predicted the rise of the "Big Nine" (Amazon, Apple, Meta, Microsoft, IBM, Baidu, Alibaba, and Tencent). The three Chinese companies are set to stay away from Europe and the US thanks to protective (or abusive, depending on your POV) regulation, but they hold a card up their sleeves. China does not have any protection against privacy, data safety, intellectual property rights, and human rights issues, and therefore, their products can be developed much faster, tested in a massive audience with no restraints. The full capabilities of Chinese AI are set to be discovered only in a dystopian scenario where the battle between West and East becomes a belligerent one.

The last afterthought is about Europe. The continent and the European Union are set to be observers of the AI wars. The absence of major players in the dispute is in part because of the much more robust safeguards regarding privacy, labor laws, welfare state, and others. But it says a lot about the failure of the world-class institutions researching the subject to have an impact in the real world. Theoretical discussions and academic-only outputs (or almost that) are far less urgent than things to help citizens now. The academy must have independence, of course, but it cannot rest in their ivory towers just to satisfy the vanity of researchers.

The intelligence to be generated artificially is only limited by the intelligence of those who create them. The Skynet danger is real, and a situation where things get out of control is no longer a 1980s movie script's sci-fi. But the immediate is more pressing than that because AI is completely under private control outside China, which makes it much harder to stop risky projects conducted by megalomaniac CEOs who can hardly be controlled by their own board members. Safety regulation measures must be deployed now. Fighting against bureaucracy and partisan fiefdoms' interests is the first major battle to be won.