29 May 2023

    AI: Chat GPT is just the entree

    Daniel Moser

    Investment Analyst

    Email Daniel
    Daniel Moser

    Investment Analyst

    Email Daniel

    The popularity of Chat GPT has seen interest in artificial intelligence explode in recent months, and companies will need to adapt to keep pace with this emerging technology. In this second of a three-part series, we look at which companies and industries will benefit from the recent leaps in AI innovation, and which will need to adapt or risk being left behind.

    “…since March, we saw a significant spike in student interest in ChatGPT. We now believe it's having an impact on our new customer growth rate.”

    These were the words from Dan Rosenweig, CEO of online education company Chegg during their latest earnings call. Chegg’s shares subsequently fell 48% in one day, sending shivers down the spine of their investors. Chegg provides homework help, digital and physical textbook rentals, online tutoring, and other student services. Rather than signing up to Chegg’s platform, students are instead opting to use ChatGPT for assistance with their homework, hindering the company’s growth.

    While AI is still in its infancy, its potential has captured the minds and imaginations of scientists, politicians, and corporations, with ChatGPT and Bard at the forefront of public consciousness.

    While AI will be a boon for many companies, there are others whose business models could be upended by the technology. But like many technological shifts – the prudent strategy can be to invest in the companies that provide the tools and services that support the growing industry – the ‘picks and shovels’. With an early lead in AI – Amazon, Google and Microsoft are well placed to play the enabler and backbone of the AI revolution.

    Large language models could transform business productivity

    Many companies have already been utilising AI for computational and quantitative tasks. JPMorgan Chase bank is using AI to automate tasks such as fraud detection and risk assessment, freeing up employees to focus on more complex tasks, Mastercard uses AI to improve fraud rates and decrease transaction decline rates, and Netflix is constantly developing new AI-powered features to improve its recommendation engine.

    While these technologies continue to evolve, the big recent innovation has been the wave of releases of Large Language Models (LLMs) such as ChatGPT. LLMs have moved AI beyond just numbers and quantitative tasks – into language, creative content creation, and replicating human interactions.

    Productivity software stands to benefit from these recent AI innovations. For example, GitHub, Microsoft’s coding platform, rolled out their AI feature named Copilot which aids developers’ code writing by offering autocomplete-style suggestions as they code.

    There has often been a talk of 10x super developers in jest, these are developers that are 10 times more productive than the average developer. GitHub CEO Thomas Dohmke strongly believes that 10x developers can become reality - “With AI at every step, we will realize the 10x developer.”

    Some business models could be upended by the technology

    Chegg is one example of an early company impacted by AI, and companies that have a large component of people doing repetitive and manual tasks are also likely to be impacted.

    Industries which previously required people with specialist skills, such as copywriting and website design, can now use AI to help complete these tasks in minutes, rather than days or weeks. Instead of using a copywriter, simple prompts can be fed into Word Copilot to generate full articles and content for publishing and advertising, and websites can be built simply by taking a photo of a hand-drawn website design and feeding this into an AI model.

    But this isn’t necessarily a negative for these industries if they can use the productivity gains on more mundane tasks, and spend more time on ‘value-add’ activities. Satya Nadella – CEO of Microsoft – believes AI could enable “every developer to focus all their creativity on the big picture: building the innovation of tomorrow and accelerating human progress, today”. The same applies to copywriting and website building.

    The companies providing picks and shovels for the data gold rush

    Perhaps the largest beneficiary from AI will be the cloud computing companies. AI is computationally intensive – it’s estimated that ChatGPT costs NZ$1m per day to operate. One conversation on ChatGPT costs only a few cents but when used by billions of people many times over this cost escalates exponentially.

    AWS, Azure and Google Cloud Platform (GCP) dominate the cloud compute market and AI has expanded their horizons significantly. Only a handful of companies ultimately have the scale and ability to provide the computing infrastructure to host AI applications. For example, both Amazon and Azure have made large investments in building their own AI specific chips for training and running AI models more efficiently – something that smaller competitors are unlikely able to replicate.

    We think AWS, Azure and GCP are best positioned to capture the lion’s share of AI workloads. This will be either through licensing their own developed foundational AI models or simply providing the computing muscle power to host AI models from other vendors, in the cloud. Despite being one of the largest software companies globally, Salesforce have joined forces with OpenAI and ChatGPT rather than build its own AI tools from scratch.

    Microsoft plans to roll out a similar AI features across their venerable Office365 and Power Platform suites. With over 350 million commercial subscribers to these suites, it is easy to imagine how lucrative this feature will be, for both Microsoft and its customers, particularly in the current environment where labour and productivity shortages are an issue.

    AI can be likened to the internet in the 90s, before smartphones, app stores, streaming and social media were on the radar of most people. Current commercial adoption remains relatively low and there are unlimited use cases. We think being invested in the picks and shovels like AWS, Azure and GCP, will offer the greatest likelihood for success regardless of which platform or AI model dominates.

    Like in the infancy of the internet, we can only guess what AI’s ultimate size, impact and application will be, but what we do know is that companies will have to adapt to harness the power of AI or risk being left behind.

    Want to learn more?

    Next month, in the final article of our series on artificial intelligence we’ll look at some of the impacts on society from the rapid rise of this new technology. If you missed our first article, we explored whether Google’s dominance will be weakened by the rise of artificial intelligence.

    If you would like to talk to someone about your investment strategy, the team at Fisher Funds are here to help. Please contact us or get in touch with your adviser.