Much has already been written on recent developments in artificial intelligence (AI) and what it means for the economy, jobs and productivity. AI caught fire in November 2022 with the release of ChatGPT, a large language model-based chatbot designed to take written language user inputs and produce a conversational written language output. ChatGPT was arguably the first chatbot – one that could converse at length and respond to successive prompts, no less – most people had ever encountered.

Unsurprisingly, ChatGPT exploded in popularity upon its debut. And given its earth-shattering potential, AI will almost certainly have wide-ranging and unforeseen effects on every thinkable industry and profession. However, the overall impact of generative AI (e.g., ChatGPT) on various professions and industries is not likely to be fully known for years to come.

For investors, examining what the spread of AI and cross-industry generative AI integration means for portfolios is an entirely different matter. With many companies already announcing multi-billion-dollar investments into generative AI, the investment implications may be far-reaching as well. Investors looking to put money to work in areas that may benefit from AI’s impact could be asking themselves: How is AI projected to significantly impact the economy?

How does AI work?

A lot of resources go into building, training and powering AI models like ChatGPT. In other words, it takes a mountain of compute power. Large language models (LLMs), or AI algorithms that leverage massive volumes of data, need enough compute power to analyze trillions of words and develop billions of interrelationships almost instantaneously. Of course, this requires significant infrastructure and hardware; to illustrate, the framework needed to train a single LLM can easily cost millions of dollars.¹

Any technological advancement that touts wide adoption is sure to bring disruption with it. This means that these technologies – ones swiftly taken up by swaths of the population – have the potential to destroy just as many industries and businesses as it helps. It may be beneficial to look for sectors and companies that are in a position to weather storms and potentially come out even stronger on the other side.

AI: Not a cure-all

Despite its massive potential, generative AI isn’t good at everything. This is because generative AI models such as ChatGPT are prone to giving incorrect responses to questions with a single right answer, such as most math problems; in practice, it’s much better at providing creative responses.² These LLMs are, after all, simply providing answers based on relationships between words in the English language and not necessarily answers based on knowledge of a fact.

But a more pressing risk is bad actors who can exploit generative AI models (like ChatGPT) to target individuals, companies and institutions. At its most basic, generative AI models can be used to write and distribute spam emails or robocalls far faster than a human could. Another potential risk that has already appeared is “prompt injections.”³ With prompt injections, hackers can write prompts to manipulate generative AI models into sharing confidential data or disclosing proprietary information.

Nevertheless, we believe that AI – especially the proliferation of generative AI – may affect investments over the medium to long term.  If you are considering learning more about AI and its potential impact on different industries, there are a few select areas within the broader sector that may benefit the most from AI’s impending widespread adoption.

Semiconductors: the power behind AI

Semiconductors are the materials that power all kinds of modern tech, including computers, cars and renewable energy. Within the context of AI, they’re critical for training and running LLMs like ChatGPT. As a result, the global semiconductor industry is poised to see growth over the coming years. The AI end market alone is estimated to generate $85 billion – $95 billion on the earnings before interest and taxes (EBIT) for semiconductor companies over the long term.⁴

Semiconductor manufacturers and designers seem set to benefit from the widespread adoption of AI. The massive investment and spending on semiconductors necessary to build future AI models underscores an enormous opportunity for leading companies. As AI models become more advanced, the investment required to train them will continue to grow, potentially benefitting semiconductor companies in turn. For example, the next iteration of ChatGPT, GPT-4, is expected to cost about $200 million to train.⁵ And that’s only for one specific LLM – there will be countless future models that will each require even more sophisticated training, and semiconductors will be critical to their success.

Early adopters: the first to benefit from AI

We see sectors and companies that are early adopters of AI as another possible investment avenue. While tech conglomerates won’t be the only pioneers, they will most likely lead the pack in terms of firmwide AI integration. Within the broader tech sector, though, the companies set to benefit the most from AI will likely include those not only capable of adapting their business models and strategies to keep pace with the changes, but those that can strategically invest in the hardware and software needed to integrate AI across businesses, too.

Companies that have the resources to adopt AI tools could be the companies that see its earliest benefits. This includes large companies that have cash on hand to invest in AI along with organizations that have access to massive compute power and terabytes of data used to train AI models. More importantly, these large, unique datasets are better positioned to craft AI models that can offer detailed business insights, which may help organizations focus on areas where there are opportunities for growth.

It's also worth noting that there will be other early adopters outside of the tech sector. Companies in the medical, advertising and entertainment industries have begun investing in AI integration methods as well. Additionally, the customer service industry is already seeing boosted productivity thanks to AI integration; research from MIT and Stanford shows that at a representative company of over 5,000 customer support agents, agents who used AI tools saw their productivity increase by 14% on average.⁶

The bottom line

Since AI is slated to disrupt dozens of industries, things will change abruptly for a large chunk of the economy. As is the case with any revolutionary technological advancement, some companies are bound to suffer as they fail to evolve and adapt to a world where humans and tech work together more closely. Considering AI’s predicted impact, a financial advisor may be able to help you augment your investing strategy with AI-driven sectors and companies that show signs of strength and long-term resiliency. Consider discussing AI with a J.P. Morgan advisor to see if it is the right fit for your portfolio and risk tolerance.