How technology became the indispensable investing partner

But what is likely less known is that the ideas behind AI stretch back more than 75 years. Through scientific breakthroughs, false starts and renewed optimism, it's a history that helps explain why AI is beginning to reshape investing.

Before ChatGPT, there was Alan Turing

In 1950, computers filled entire rooms and were built to calculate, not think. Few imagined they would one day write essays, discover medicines or help make investment decisions. Alan Turing did.

That year, he published a paper titled Computing Machinery and Intelligence. In it, Turing posed a question that many scientists of the day dismissed as fanciful: could a machine think? More importantly, he argued that intelligence might not be unique to humans, but something that could be recreated.

It was an extraordinary idea. The only problem was that the technology to test it didn't yet exist.

The first AI boom

Six years later, a small group of researchers gathered at Dartmouth College in the United States for a workshop that would give birth to the field of artificial intelligence. Their ambition was to discover whether abilities such as learning, reasoning and language could be described so precisely that a machine could simulate them.

The organisers believed the problem might be substantially solved over the course of a single summer. Instead, it became one of the longest scientific pursuits of the modern age.

Throughout the 1960s, government funding fuelled research into machines that could solve logic puzzles, translate simple text and even play chess. Then, in 1966, MIT computer scientist Joseph Weizenbaum unveiled what is now regarded as the world's first chatbot. Although little more than scripted responses, many users found the conversations held with ELIZA remarkably convincing.

For a moment, mass adoption of these intelligent machines seemed like it was just around the corner.

The long winter

Outside research laboratories, artificial intelligence was already taking on a life of its own.

Stanley Kubrick's 2001: A Space Odyssey introduced audiences to HAL 9000 in 1968, a computer capable of conversation, reasoning and independent decision-making. Two decades later, Star Trek: The Next Generation imagined a very different future through Lieutenant Commander Data, an android striving to understand humanity while demonstrating remarkable intelligence and judgement. Popular culture imagined intelligent machines long before science could build them.

The problem for these visionaries was that real life was too complicated to capture with a book of rules. Every new exception required another instruction, and every new instruction created countless more possibilities for a computer to consider.

By the early 1970s, governments began questioning whether decades of investment had produced the promised breakthroughs. In Britain, mathematician James Lighthill concluded that AI systems worked well in laboratories but struggled in the complexity of the real world. Funding was cut, research programmes closed and the field entered its first AI Winter.

In the 30 years that followed, funding remained scarce, but researchers didn’t give up. Over time, those researchers also changed their approach. Instead of trying to tell machines everything they needed to know, they thought that machines could learn from experience in the same way that humans do.

The idea remained largely theoretical because computing power and data were still in short supply.

By the turn of the century, the world had changed. The internet had transformed information into something abundant, and computing power was advancing quickly. And nowhere was this more apparent than in, of all places, a chess match.

The computer beats the grandmaster

In 1996, IBM's Deep Blue defeated reigning world chess champion Garry Kasparov, becoming the first computer to beat a reigning world champion under standard tournament conditions. For decades, chess had been regarded as one of humanity's ultimate intellectual challenges, making the victory a landmark moment for artificial intelligence.

Deep Blue's victory demonstrated the power, and the limitations, of brute-force computation. The machine could analyse around 200 million positions every second, but it couldn't learn or apply its abilities beyond chess. That meant the next breakthroughs in the AI story would come from teaching machines to learn from data rather than programming every rule.

When machines started learning

Innovation accelerated in the early 2010s. Apple's Siri brought voice assistants to consumers, while IBM's Watson demonstrated that computers could answer questions posed in natural language by defeating two champions on the television quiz show Jeopardy!

But the biggest breakthrough came in 2012, when a team from the University of Toronto entered a deep learning model called AlexNet into the ImageNet challenge. Rather than relying on written rules, AlexNet learned patterns from enormous quantities of data, identifying objects in photographs far more accurately than every other system in the competition.

Five years later, a group of Google researchers delivered another breakthrough. They published a paper titled Attention Is All You Need, and with it, introduced the Transformer architecture. For the first time, larger models consistently became better models. Give them more data and more computing power, and their capabilities kept growing.

While Siri and Watson showcased what AI could do, AlexNet and the Transformer architecture changed what AI could become. The first proved machines could learn from data; the second showed that learning could scale.

From the laboratory to the portfolio

Generative AI entered the mainstream in late 2022 with the launch of ChatGPT. Suddenly, anyone could ask a computer to write, explain, summarise or reason using everyday language. Claude, Gemini and a wave of other AI assistants soon followed.

If AlexNet and the Transformer changed the trajectory of artificial intelligence, ChatGPT changed public perception of it. What had long been confined to researchers and technology companies suddenly became available to millions. Suddenly, anyone could experience modern AI by completing tasks in minutes that once took hours.

Investing has proved no different. Until recently, professional investors held an advantage through superior data, computing power and proprietary models. Today, many of those capabilities are available to individual investors.

Today, investors can ask AI to compare ten years of annual reports, identify risks or explain how macroeconomic changes affect an industry in minutes. Some have even created whole models and run their own portfolio reviews in Claude.

But like any investment tool, artificial intelligence is only as useful as the person using it. It won't replace judgement or experience, but it can dramatically reduce the time it takes to understand an investment opportunity and allow investors to focus on higher-value decisions.

The next frontier

The first generation of AI helped investors process information more efficiently. The next generation is beginning to help them discover opportunities they may never have considered in the first place.

Rather than starting with a fixed investment philosophy, generative reinforcement learning explores thousands of combinations of company fundamentals, technical indicators and macroeconomic signals, rewarding those that consistently identify opportunities and discarding those that don't. As market conditions evolve, so too can the signals the model relies upon.

That is the thinking behind the forthcoming VanEck Dynamic International Equity ETF (GOAT)¹. Each month, its engine utilises the insights of more than 16,000 investment signals and applies them to more than 1,200 companies. Based on what it learns, the model selects 150 companies and weights them according to their assessed probability of outperforming the benchmark. The result is an investment process designed to evolve alongside changing market conditions, rather than remain fixed to a single style or philosophy.

From thought experiment to reality

History suggests that the most influential technologies rarely arrive fully formed. They improve gradually, find new applications and, over time, become indispensable. When Alan Turing asked whether machines could think, he was really asking whether intelligence itself could be engineered. 75 years later, that idea is no longer confined to university laboratories. It is beginning to influence the way we research, work and now, invest. The next chapter of that story is only just beginning.

_{1 - The new strategy will be available from Monday 20 July 2026, when GOAT begins tracking the Akros Enhanced World ex Australia Index.} 

Key Risks: 

An investment in GOAT carries risks associated with: ASX trading time differences, financial markets generally, individual company management, industry sectors, foreign currency, country or sector concentration, political, regulatory and tax risks, fund operations and tracking an index. See the PDS and TMD for more details.

GOAT is likely to be appropriate for a consumer who is seeking capital growth, is intending to use the product as a major, core, minor or satellite allocation within a portfolio, has an investment timeframe of at least 5 years, and has a high risk/return profile.