TECH NEWS – According to one artificial intelligence company, it’s not really worth praising human intelligence…

Jan Szilagyi, co-founder and CEO of software company Reflexivity, spoke on CNBC’s Squawk Box about how leading hedge funds are using AI to complement their existing tools, and he says the technology has already made several correct stock market predictions this year. Relfexivity is developing large language model (LLM) overlay analytics engines and believes that with the right computing power and training, AI can beat human intelligence when it comes to trading stocks.

Reflexivity’s ultimate goal is to build an autonomous investment analyst, which is why the company offers technologies that allow investors to aggregate data from multiple sources on a single platform to simplify decision-making. By posing questions to the engine, users can trust it to find the data and then perform the analysis, which takes two minutes instead of two hours. Reflexivity’s current AI-powered trading systems are designed to find parallels from the past to give a sense of what future price patterns might look like. These parallels are based on 12 or 15 past episodes to select securities that match our search in terms of the current environment or a set of economic or other parameters that may cause certain stocks to offer superior returns to others.

Szilagyi’s clients are primarily the best hedge funds, which use it as a kind of intelligent overlay on top of the various data sources they have access to. AI has hit some stock market trends this year, but missed others. Szilagyi explained that if you look at the market as a whole, for example, it was very correct in predicting the July top, it was very correct in predicting the false rally that occurred immediately after. After the top, the MI was again correct in signaling the bottom. But its software turned bullish (a term that in stock market parlance means rising stock prices) too soon before the Federal Reserve’s latest meeting, after which the bank scheduled two rate cuts by 2025 instead of four. However, Szilagyi does not believe the model is wrong here either, as its outputs are always probabilistic assessments. According to him, these judgments essentially mean that the model predicts with 70% certainty that something will happen, but there is still a 30% chance that it will not happen. He says it is hard to say that there is anything special about human intelligence that should make us better at trading. He believes that if we are able to put so much computing power into the system, and intelligence is constantly improving, then at some point it is not crazy to think that systems will be better at trading than humans.

He concluded that we are still about 5-10 years away from replacing humans, and he believes that AI will not completely replace us because some sectors, such as private markets, will be harder for AI to target due to the relative lack of data needed for training.

Source: WCCFTech