Are you a leader feeling AI stress?
Just ”keep calm and carry on.” You are normal.
That old British saying works well here.
You are not alone in being overwhelmed, impressed, taken aback and maybe also disgusted. Just keep calm and carry on. We need more calm and reasonable people like you.
We don’t want to get sucked into the most extreme parts of this hype cycle, but of course we welcome human-friendly tech evolution. We want the best of it.
I’d like to show you that you are not alone.
1) Is this AI revolution really as big as people are saying?
Money-wise, yes. For now.
About a month ago, it was reported that OpenAI is -again- hunting for more money. This time they want to get their hands on 100 billion USD. Other LLM giants are doing the same.
This ’cash injection’ (it’s a bit bigger than that, wouldn’t you agree?) is partly meant to help the company sustain their running costs.
Read that again. Those are some high running costs.
If they get that money, it’s been reported that OpenAI will be valued at 750-830 billion USD. That is comparable to the GDPs of countries like Belgium, Ireland, Switzerland or Saudi Arabia.
If you think this can’t go on, you’re not alone. A lot of people are raising flags here. If you think this is bonkers and you smell a rude awakening or some sort of crash coming, you are not alone.
…and before you say that AI tech actually delivers, please read this article from The Guardian.
”The huge gap between capital investment in infrastructure and end user license revenue from AI software is not sustainable.”
”…even if the technology delivers on its promise, many, if not most of the current players are bound to fail.”
You were already thinking that the investment rate sounds crazy, right?
You were right. Keep calm, carry on. Let’s stay calm together.
2) Where is all that electricity supposed to come from?
You have been thinking about this, right? You are not alone.
In Sweden for example, companies need to apply before they can start something that consumes very large amounts of electricity.
Data centres need huge amounts of electricity.
Swedish authorities currently have a pile of applications that they can’t approve.
Why? Because in order to to approve all the current ’I want to build a data centre’ applications, Sweden would need to have two or three more nuclear plants. Today. Just to approve the existing applications.
3) AI-based or AI-augmented products everywhere are being touted as ’fantastic’.
There are a lot of very impressive things coming out, but one clear sign of a hype phase is that rubbish in shiny colours is also thrown out into the market.
This is also true for AI-related things. Even big tech are doing it.
Sometimes, it’s the use cases that are rubbish, sometimes it’s the products.
Perhaps you’re afraid to speak up and sound too negative? You are not alone.
Here are a couple of examples. I’m sure you have others.
Example A:
I recently came across a Google training video where AI was presented as the solution to this tricky use case: ”How to find out why a door handle is a bit loose.”
The AI instructor used a phone to record herself wiggling the door handle and then asked the AI for help.
The AI’s answer: Tighten the screws.
The instructor then asked what type of screwdriver to use.
The AI’s answer: A Phillips-head or a flat-head, depending on the type of screws.
Thank you, AI!
Example B:
There are a number of products available that supposedly help you create great-looking (’stunning’) Powerpoint presentations.
First of all, let’s remember that Powerpoint is a tool to present things, not the final solution to anything. Some of those tools are difficult to use, but it actually gets worse than that.
Some tools may ask you for directions and then still only give you images than you have to paste into Powerpoint slides yourself. The images can’t be edited. Not very useful.
After that, they offer you to ’iterate’ (spend more and more precious time) in order to get something you may be able to use.
Other tools can actually create real Powerpoint slides, but those slides are sometimes worse than you’ve seen in a long time. After that, it’s up to you to direct the AI to ’refine’ them, step by step.
…and then there are of course tools that can actually do good things. Nice work!
How much time and money should we spend on making fancier slides?
Shouldn’t we be somewhere else by now, doing real work that benefits our customers?
Keep calm, carry on. Let’s stay calm together.
Take care everyone,
Björn