Everyone’s falling over themselves trying to get a grip on artificial intelligence – the cause of, and solution to, all of life’s problems. Two years ago, while most of us were laughing at AI-generated photos of people with extra fingers, scientists and tech CEOs were debating whether AI posed an existential risk to human civilization. Now our Oscar-nominated movies are all Botoxed up with generative video, and the AI debate is over: We’re hooped.
Let’s dial down the angst for a moment. Much of what we ham-handedly call “artificial intelligence” is just brute-force computation done at increasingly faster speeds. As always, it’s the humans using these technologies that we need to pay attention to. So let’s see who’s out there using AI in the crypto markets, and how their combined efforts are shaking up the industry as a whole.
What Is Artificial Intelligence?
Before we get to our players, we need to nail down our central theme. When we talk about human intelligence, we’re talking about our ability to learn and solve problems through reason – as opposed to our more intuitive faculties. Artificial intelligence, then, is any human-made thing that exhibits the same ability.
Which means we’ve been doing AI since ancient times. Those sacred Egyptian statues may not have been particularly effective AI agents, but people believed they were capable of thought. More complex automatons would follow; the Antikythera mechanism from Hellenistic Greece is considered the world’s first known analog computer.
Long story short again, the AI tools we use now are just the latest versions of golems and ENIACs. So when people think about AI in 2025, chances are they’re really thinking about today’s No. 1 tool: large language models, or LLMs. These are the learning models behind ChatGPT and all those other generative AI programs taking over the crypto news.
What Do LLMs Have to Do With Crypto?
As much as we’ll let them. Large language models are being used right now to develop, operate and maintain blockchains; beyond simply churning out code, LLMs can parse tremendous amounts of data, finding and flagging anything of concern in real time. Everything from individual transaction records to entire crypto markets can be analyzed using these AI tools.
Which is where we join our central cast. As pointed out in a February 26 MIT Sloan article by Betsy Vereckey, the crypto market has become much more concentrated in recent years within a small number of exchanges, with Binance at the very top of the mountain. In theory, you can still create your own exchange and your own token anytime you want, but if you want to use the same LLMs the big guys are using, you’d better have tons of cash at your disposal.
To be fair, the consolidation of the crypto industry began even before the latest AI boom, with the November 2022 failure of FTX playing a large role. And we’re not saying that Binance or any other exchange is necessarily wrong to use LLMs and AI in general to increase their market share. We’re simply reporting the shape this leaves the industry in: top-heavy.
What Does This Mean For My Portfolio?
Buckle up. With so much power in the hands of so few exchanges, the downfall of any one titan can be catastrophic – just look at what happened in November 2022 when FTX went belly-up. Volatility is already baked into the crypto market, so if you’ve been investing for a while, you know the drill: Diversify, and risk smaller amounts on these speculative assets.
The larger problem is with the market itself. Cryptocurrency is supposed to be decentralized, yet here we are with Binance, a privately-owned centralized exchange (CEX), controlling 34.7% of the market as of December 2024 according to CoinGecko. True, Crypto.com (11.2%) has eaten into Binance’s lead somewhat, and Upbit (9.8%) is making noise again, but is this the vision Satoshi Nakamoto had in mind when he/they launched Bitcoin in 2009?
