Artificial Apocalypse

AI is inescapable currently. Quite literally every app is introducing some sort of AI into it’s main sections. My own website admin panel now has it through WordPress which seems to have included it “out of the box” – I have no clue to which service it pings or who is footing the bill for the use of that AI (aside from the environment of course but let’s not get bogged down there). Nor do I have any clear idea of how to turn it off, will need to hunt that down later. Regardless to say this bubble is inflating fast, and I call it a bubble for several reasons.

The recent study conducted by MIT sent shockwaves rippling across the AI industry stating 95% of projects started with AI fail. My main thought is much more feelings & vibes based. From what I’ve seen over the last few years with web 3.0 & NFTs, any technology that tells you it’s going to revolutionize things and to get in on the ground floor sounds a little too speculative and riding a hype wave. Anything that truly moves the needle usually does it gradually & quietly so that by the time you noticed it’s too late to join the bandwagon so that all the money is consolidated already.

The real issue will always be the arms racing of pushing unethical uses of technology without the proper diligence and precautions taken. The focus should be if we do inadvertently stumble over AGI we haven’t fed it the dogma of zero sum & profiteering but rather empathy and abundant flourishing. There is a plethora of excellent tasks that lean very neatly into the use of LLMs and the new wave of agentic models. To do so without recognition of the human and environmental impacts widespread adoption of these tools has had in the short term, would be negligent at minimum. The long term effects of outsourcing our cognition to algorithms is not yet know, that is small fry when compared to the full offloading of our psyche and social interactions to a chatbot on steroids mixed with keratin & intellectualism. Thinking AI will automatically 10x everything it touches, alleviating the need for any skilled artists, coders, copy writers, designers, marketers, novelists, researchers, executive assistants, teachers, lawyers and whatever other jobs that have been claimed can be replicated by one chat gpt5 model; shows a grave misunderstanding of all the soft skills needed to accomplish anything in a small to mid-sized business let alone anything at scale. And whilst this current point is the worst it will be going forward, the rate of improvement is starting to plateau significantly, so much so that the bidding market dipped significantly wth the release of gpt5.

My hope is that the open sauce market of models doesn’t stagnate too far behind the so called “frontier” models so that the average user can utilise the services & power of the LLMs without being beholden to a few oligarchic corporations.