Lower Manhattan’s trading floor is quieter than usual just before the opening bell. The shouting, the chaotic energy that once characterized Wall Street, has subsided into something more contained, but screens still glow in rows and numbers still flicker. Traders watch real-time model updates while seated with dashboards and coffee cups. It appears that the actual action is taking place elsewhere. within devices.
Algorithms were tools ten years ago. They feel more like participants now. According to rough estimates, automated or AI-assisted trading accounts for about 70% of market activity. It wasn’t a sudden change. It first appeared in high-frequency trades, and later it surfaced when machine learning models subtly outperformed human analysts on specific tasks. These days, it’s difficult to find any area of finance that isn’t affected.
| Category | Details |
|---|---|
| Industry | Financial Markets / Trading |
| Core Technology | AI, Machine Learning, Algorithmic Trading |
| Trade Volume | ~70% of trades involve algorithms |
| Key Players | Goldman Sachs, JPMorgan Chase, Blackstone |
| AI Market Size | ~$11.2B (2024), projected growth ahead |
| Trading Types | High-frequency, quantitative, automated |
| Key Advantage | Speed, data processing, pattern detection |
| Main Risk | Market volatility, flash crashes |
| Human Role | Oversight, strategy, risk management |
| Reference | Built In – AI Trading |
Not only has speed changed, but so has interpretation. These days, AI systems do more than just respond to price changes. They scan news headlines, read earnings reports, and even use tone analysis to determine, in a matter of milliseconds, whether a sentence sounds cautious or optimistic. Machines may be able to “read” the market more thoroughly than any one person could in the past.
This change is evident in subtle ways within companies such as JPMorgan Chase and Goldman Sachs. In order to create summaries, models, and even draft investment theses, junior analysts who used to spend their late nights buried in spreadsheets now rely on AI tools. The artwork has evolved, but it hasn’t vanished. There’s a feeling that the entry-level grind—the practice of repeatedly proving oneself—is disappearing.
This efficiency appears to be welcomed by investors. Now more than ever, speed is crucial. An AI system identified a specialized supply-chain trend days before rivals did, according to a portfolio manager. A profitable trade resulted from that early signal. Little advantage, large reward. Businesses are pursuing these moments, which are repeated in thousands of decisions.
However, there are peculiarities in the system. Algorithms don’t think like people do. They react to correlations, patterns, and probabilities; occasionally, they amplify signals that are not entirely understood. When several systems respond simultaneously, the outcome may resemble a feedback loop rather than a market. Prices fluctuate quickly. Then more quickly. Sometimes, too quickly.
The 2010 “flash crash” is still fresh in people’s minds. Markets fell precipitously in a matter of minutes before rising again, leaving traders frantically trying to figure out what had happened. Such incidents are uncommon, but they pose an unspoken question: if trading is dominated by machines, then who is in charge when things go wrong?
A more profound cultural change is also taking place. Data and instinct—gut feelings developed over years of experience—have always been the driving forces behind Wall Street. AI poses balancing challenges. Instincts are absent from models. They possess training data. They become consistent as a result, but they may also become blind to things that don’t fit historical trends.
AI is being used at companies like Blackstone not only for trading but also for assessing entire businesses, including forecasting performance, identifying risks, and even making strategic recommendations. It’s an impressive and a little unnerving expansion of scope. There is a perception that decision-making is becoming less centralized and more distributed.
Despite all of its benefits, artificial intelligence creates new risks. Overfitting of models is possible. Incomplete data is possible. Additionally, markets change by nature. What was effective yesterday might not be effective tomorrow. It’s still unclear if AI systems will magnify truly unpredictable events or if they can adjust quickly enough.
The issue of access is another. Big businesses can hire teams of engineers and data scientists because they can afford state-of-the-art technology. Conversely, smaller players might find it difficult to keep up. Over time, that disparity might grow, consolidating power in a smaller number of hands. It’s a possibility that lurks in the background but isn’t publicly discussed.
It’s difficult to ignore how rapidly the culture has changed as you watch this develop. Now, it almost seems nostalgic to see the fast-talking trader making snap decisions based only on intuition. It is replaced by something more subdued and deliberate. Although decisions are still made, more and more of them are being shaped, filtered, and carried out by machines that are faster than humans.
However, the human component hasn’t completely vanished. The models are still constructed by someone. What data is important is still decided by someone. When something doesn’t seem right, someone still pulls the plug. For the time being.
Wall Street seems to be in a transitional stage, striking a balance between machine precision and human judgment. It’s unclear if that balance will hold or if it will tip entirely in one direction. One thing is certain, though: the market is no longer moving at a human pace. It hasn’t in a long time.
