Artificial intelligence is best viewed as a business accelerator — something that adds new “smart features” to existing products, and boosts tech companies’ revenue?
Is it a new disruptive platform that has the potential to shift the balance of power within the tech industry.
Since ChatGPT exploded onto the scene, this question has been lingering in the air. Many software products will benefit from generative AI, which can create an essay on demand or a picture at your command. What if the software you use could predict your next need, and make a creative suggestion for it?
The technology, by offering a new way to interact with computers, could become its own platform. OpenAI’s large-scale language models are already used by many applications to enhance AI capabilities. In the past, technology platforms like this one have been very consolidated — whether it’s software for laptop computers (Windows and MacOS), smartphones or browsers. How many AI “foundational models” will be supported by the market?
For now, investors are focusing on AI as a catalyst. The questions asked during this week’s Microsoft, Alphabet, and Meta earnings calls were about how generative AI enhancements would translate into increased revenue.
Microsoft’s announcement last week that it would charge $30 a month for the AI features they plan to add to their productivity apps was the most confident statement yet regarding the upcoming generative AI wave.
If large language models are to become the core technology for many future products and services, then control of an advanced LLM could be a requirement to enter tech’s top leagues. Microsoft and Google, with their close ties to OpenAI are far ahead of the curve in terms of developing this technology. Can they match them in their race to create ever-larger models of AI — and what would it mean if others couldn’t?
Apple, Amazon, and Meta tend to view machine learning as a key ingredient in their products. But it is still just an ingredient. It’s something that can be used to improve the recommendation system or to provide analytical tools for advertisers. Now they are rushing to make AI a central part of their operations.
Microsoft, Amazon’s closest competitor in cloud computing, seems to have caught Amazon Web Services off-guard when it comes running AI services. Microsoft has predicted that AI will add over 2 percentage points to its cloud computing business in the first quarter. AWS has been working hard to catch up. This week, it announced several developments including a broader range of LLMs offered by other companies.
Meta’s 3.9 billion users, who download the app at least monthly, represent a huge potential audience. Mark Zuckerberg, the chief executive of Facebook, gave us a sneak peek at what we can expect. He said that a variety of AI-powered agents, avatars and assistants will soon populate their services. They’ll help users create better content and increase their interaction with chatbots.
Meta has developed a set of open-source models for languages called Llama to support these ambitions. Amazon has its own language models called Titan. Last week, it was reported that Apple also entered the race.
Zuckerberg said to Meta’s Investors this week that there wouldn’t be “one AI that people interacted with”, but that different tools would be preferred by creators for different purposes. Adam Selipsky, AWS’s chief executive officer, made the same prediction in a Financial Times interview: There won’t just be one model that will rule them all.
It’s too early to predict the outcome. It was widely believed in the early days of the internet that there would be a variety of competing services, and specialised search engines would fill different niches on the market. In the end, one general-purpose service dominated. OpenAI is betting that AI will also benefit from building larger language models.
If Wall Street’s optimistic forecasts prove correct, it won’t be long until the first fruits of the generative AI revolution start to appear in the earnings reports of tech companies. It will be a long time before the full impact of generative AI on tech companies is known.