Hands typing on a keyboard with AI magic circling to help make things work

AI isn’t going to lead to large-scale enterprise layoffs – yet!

Glenn’s view on three ways AI isn’t ready for the Enterprise in 2024, or maybe even 2025.

  1. Copyright and IP protection aren’t in place.
    Large corporates are quite rightly very protective over their Intellectual Property and anything else that keeps them unique in the market. Therefore, I don’t see corporate legal teams signing off on using AI until they have legal protection over the source material that seeded the AI. They cannot take the risk of building their products on top of something that might later expose them to legal action or require them to open-source their IP.

  2. Trust in the tooling isn’t in place.
    It may be fun to have an online dialogue with ChatGPT or summarise a document with Bard, but for an enterprise to buy a tool, use it in anger, and then fire huge swaths of the workforce in exchange for said tool requires a leap of faith. Building this trust will not come overnight, so there is a lead time from installation to mainstream that needs to occur.

  3. Compute power isn’t in place.
    Silicone is still expensive because there isn’t enough of it around, and the type needed for AI is the most expensive to buy and operate. Until the supply becomes freer and the bang per buck (or kWh) improves, it creates challenges over the economics and, more importantly, the predictability of your supply to run and grow your business. 

Will these reasons last forever? No, but I don’t see them falling in the short term. Over time, they will be chipped away at, and some lower-level roles will pass to AI because enterprises know they can roll back to humans easily and cheaply. Deep specialisms and technical skills are the areas that enterprises need to have confidence that AI is competent before they lose the organisational knowledge humans maintain.

If you would like to have a chat over a coffee with Glenn to explore this topic, please get in touching.