Veiled in morning mist
Some point in time observations about GenAI. What is it, how much is being invested, where is the value now and which industries face disruption?
Time for a bit of a stocktake.
GenAI based on LLMs has established itself as a new technology tool.
The underpinning technology is developing rapidly. Although there is nervousness that it may be reaching a plateau.
At the moment, GenAI needs heavy upfront investment and has high operating costs. There is a theme that this changes the business model for SaaS and software in general. That depends on how far and how fast the costs fall. There are signs that fall is starting to happen.
Investment
Heavy investment is happening. The Economist estimates spending on data centres alone could total $1.4 trillion between 2024 and 2027. 46% of software investment is in AI.
Valuations of public and private companies in the US indicate very high expectations for returns on that investment.
Largest investments are by established big tech companies - Microsoft, Alphabet, Meta, Apple, Amazon. Some of the "data" is overstated because the investees spend a chunk of the money buying services from the investors, for example Microsoft and OpenAI.
Hard to see if this is competition or an oligarchy forming. If it's the former, Alphabet seems to be edging ahead but that could change quickly.
China and other parts of Asia show a similar pattern of activity. Investment numbers are obscured because much resources are directed and provided through the state.
But the Chinese economy is struggling. This is a predictable consequence of reverting to a more central state model and restraining market freedoms. How strong will China really be in AI?
Europe started from a bad place and is falling further behind rapidly. France is a partial exception but its position looks brittle.
Where is the value?
GenAI offers tremendous potential but high value use cases have not yet been proven.
We can see some early value emerging in businesses:
Developing software more efficiently. Lots of publicity on this but devs I talk with indicate it's still early days....
"The jury is still out on whether Copilot is useful enough to justify the cost.” — Chevron CIO
Marketing. You have seen this whether you like it or not. Is this effective or just flooding every channel? Walmart example shows how it can deliver practical value.
“We’ve used multiple large language models to accurately create or improve over 850 million pieces of data in the catalog. Without the use of generative AI, this work would have required nearly 100 times the current headcount to complete in the same amount of time.” — Doug McMillon, President & CEO Walmart
Customer support. Most likely you have seen this in action as well. Klarna is the most public example. They now claim to have stopped hiring and seen a 23% drop in headcount. Appears real but no flood of fast followers yet.
Some back office task automation. Very limited evidence for this.
Having said that, some common personal use cases are closely linked to task automation.
Reading and summarising including voice/ audio for those who don't like big blocks of text.
Meeting notes and summaries.
Messing around with images and audio in various forms.
All about cost reduction/ efficiency. No evidence of huge growth drivers.
Disruption ahead
Current startup activity and the focus of much investment tells us there are some industries where GenAI is expected to drive major disruption:
Healthcare
Education
Professional services - legal and consulting especially
Media - not sure about this one. Already massively disrupted so it feels like just an extension of the same complaints we have been hearing for years.
Barriers to change are high in these sectors. Regulation, custom and practice, vested interests all suggest real change will be slow and painful.
The opportunities are enormous. Unlocking the puzzle of services productivity would generate decades of strong economic growth.
This is not just about money. Services productivity is essential to ensure the future sustainability of public services, especially health and social care.
What’s next?
It feels like the big changes are all in front of us. I will start the New Year with some observations about what’s next. Meanwhile, enjoy the break.
Thanks for reading.