Every trader is familiar with the word “edge”. You need it to be successful. Edge can take many forms. You may have a better understanding of the market than others; your analysis might be sharper than conventional wisdom. Or you could be so adept at reading price charts that you are able to recognize patterns and know the exact moment to buy or sell.

Edge is difficult to find. The majority of people lack it. According to studies, more than 80% day traders lose money.

What will happen to the traders’ edge if AI floods the crypto markets? Sheraz Ahmed, Managing Director at Storm (a Blockchain consultancy) recently said to me that AI will be the enemy of those who are stubborn. If a trader doesn’t want to use [AI] or wants to do it manually, then they will be left behind.

This is no trivial matter. We may not like it, but the main use of cryptocurrencies — by far — is the simple buying and selling, which often takes the form of scalping.

Jeff Wilser: 10 Ways Cryptography and AI can make each other better (or maybe worse)

The AI trading revolution may have already started. According to Ian Allison of CoinDesk, who reported on the matter, Chain of Demand is an investment data analytics company that works with blue-chip institutions such as Bloomberg. They have developed a tool which leverages ChatGPT in order to analyze bitcoin prices fluctuations. OkX, a cryptocurrency exchange, recently published an article on the “9 best trading bots for 2023,” including “Dollar cost averaging bots, options bots and HODL Bots.” Crypto Twitter is flooded with ads for AI bots , such as, “This Crypto Ai trading bot makes $100 a day!”

AI trading is appealing to many because you can use it and go on with your day. Daisy, a crypto-education influencer known as Crypto Emissary, says that “a lot of people have trouble knowing when to sell.” Daisy is now working with two AI trading platforms, Otto Bought and DeFi Trading Club. The DeFi Trading Club and other AI trading projects, such as Otto Bought (get it?) can help you with technical analysis and macro studies, like evaluating the tokenomics structure of a crypto startup.

Daisy says, “This will change the way people make money.” “People don’t need to sit at the computer anymore.” AI bots will do it for us. You can cash out your money to do real life things.

Some traders are less confident. Christopher Inks of TexasWest Capital, the trading group’s founder, says, “Realistically we aren’t there yet.” He is aware that the banks and large trading desks are claiming to use AI. However, he says, “we do not have enough information about how AI impacts their trading and how they invest and allocate.”

What about the AI trading bots, then? Inks believes that these are the successors of algorithmic trading robots, which were long a favorite among novice traders looking for easy wins. Inks says that “everyone and their grandmother swears by their bot’s 90% win rate strategy.” “They don’t.”

There are no silver bullets

Even if AI bots aren’t a panacea, it is easy to imagine an era where they won’t be beaten. Here, I have first-hand knowledge. Over the last two years I have tried day trading U.S. stock (not cryptocurrency). You should create a set rules, a “playbook”, and strictly follow them to avoid trading on emotion.

This philosophy is adopted by many day traders. Adrian Zdunczyk is the founder of The Birb nest, a trading group. He says: “Our edge comes by being robotic.” “Our edge is derived from the use of reliable steps with predictive value.”

Zdunczyk does not believe that AI bots’ advantage for traders will continue as the technology becomes mainstream.

That’s exactly what I did. I stuck to my plan and only took trades that met all the criteria.

What’s the point? It’s hard. It’s hard.

I was robotic with my rules. But the truth is, I didn’t have the confidence to know that my rules would yield profits if I followed them to the letter. It’s hard to tell if a trade worked because you followed your playbook correctly or just happened to be lucky. If you lose a trade, you may have followed your strategy perfectly but the market moved against you. This is an acceptable loss. Even the most successful traders will lose money. Zdunczyk at Birb nest says that although he is wrong 70% of the time, his 30% winning trades are so profitable, he still comes out on top.

Read more: Crypto miner Hive Blockchain touts privacy of AI models running on its GPUs

It is important to note that, whereas the traditional bots were only able to be programmed in a way where they could follow specific instructions, like those I used — buy, if Conditions A, B, C, and D are true — the next generation of AI may bring something new: Pattern Recognition.

Here’s a quick analogy. Remember the “China spy ballon” that hovered above Alaska? AI was used to find its trajectory. Synthetaic is a startup that used AI to process a sea of satellite images. The AI was able to process all of the images in just a few seconds, whereas it would have taken an entire human lifetime to go through them. A human trader would have to click through hundreds of thousands of cryptocurrency charts to find the exact configuration that matches his tradebook. Once AIs become better at pattern recognition they will be able to do this instantly.

Some traders swear by chart patterns like ” Head and Shoulders” which appears as a small hill to the left followed by a large hill in the center, then another smaller hill on the right. This is a bearish pattern and could be a signal to sell. You’d also consider many other factors. The AI bot would scour all cryptocurrencies — not only BTC/USD, but also more exotic pairs like SOL/BTC – at every second, around the clock. It would be looking for Head & Shoulders patterns and selling if they met a set of conditions.

Zdunczyk says that AI helps to spot more complex patterns in nature. It gives you greater accuracy. Or maybe the AI will laugh at the Head and Shoulder pattern and come up with its own way to find an edge. In chess, AI developed strategies for winning that humans had never used.

Perhaps crypto AI tools such as Otto Bought or Verox can already eat up market inefficiencies, or perhaps they aren’t. It’s not a relevant issue, I would say. It’s just a matter time before they are. Zdunczyk has been using ChatGPT for scripts to create trading bots.

He sees it through a broader lens, and as part of an ongoing trend. Traders used to have to be physically on the trading floor in order to shout out their buy and sell orders. After the shift to digital trading, high-frequency trading and free apps like Robin Hood, we saw a rise in these technologies. The market is constantly evolving.

You will need AI at some point or you will be stuck with bows and arrows, while other armies use machine guns. Zdunczyk predicts that “the advantage will arbitrage away.”

Once everyone uses AI, what happens?

Those who create the fastest, most intelligent, and beefiest AI systems may be the only ones to win. If you use it, and everyone else is doing so, you may not be able to find any advantage.

Ben Schiller is the editor.