One may read articles proclaiming the benefits or conversely the tragedies of AI. I lean towards the tragic end of this scale, which comes from my experience and exposure to emerging technologies, allowing me to see how this will be another “.com” boom - yes there will be some good bits, but the majority of AI will amount to a waste of our precious resources.

A person close to me, whom I respect greatly from both a technical and business angle, recently said: “AI will be like email, yeah the first people to get email had a competitive advantage, but anyone who eventually got email didn’t go out of business.” I think this perfectly reflects where the world is; basically in a bubble. Companies all around (including ones I work for) are racing into “AI” in fear, rather than interest in the technology. There are those who should know better, looking at you big-tech, there are also some genuine startups and real uses of AI, or should I say LLMs that are coming to the fore.

This is all to say that I don’t want to be one of those naysayers. However, my position and experience can only conclude that most businesses are in a total AI haze, jumping onto a bandwagon that is supported by a cyclical monetary system.

I also have an issue with the use of “AI” to describe what is basically a statistical model; LLMs are not the future, they will form part of it, but if AI is to truly take off we must be cognisant of all aspects of AI, not pigeonholing ourselves into LLMs just because someone has figured out you can charge for API access to them (at a loss I might add).

I’ve blogged before about feeling like my professional role as a Platform Engineer will be to clear all the robo-barf from the ship, which I think is slowly coming true. Some of my fear is tied up up in the realisation I might not even understand what is in the barf. Conversely, keeping my skills up to date makes me feel guilty for participating in the destruction of our planet via data centres.

—

What can be done then? A good place to start is thinking for ourselves. Look back to when Google wasn’t a verb, or Wikipedia could still be practicably printed, many cried “this will ruin people’s ability to think” - it probably did, but it also opened up a world where we could be better informed and better educated. Just because people choose not to think can’t really be blamed on Wikipedia, that is actually the fault of social media.

What can AI do? AI can help us with our day-to-day tasks, even be integrated into our personal lives, but what we must not let it take over everything. Please do review that PR twice, don’t “vibe-code” a product to production, write that post from your own mind, don’t ask ChatGPT. Conversely, let ChatGPT respond to an idiotic letter from your GP or HMRC asking for the same information for the fourth time, do let AI solve protein folding problems we’ve yet to crack.

Embrace AI for what it is good at and allow humans to thrive at what we are good at: humanity. Just pray that the big-tech money pile is large enough to result in a soft landing for the wider market, when the bubble does finally pop.