Unveiling the AI Hype
What comes after the Peak
We see currently an overwhelming AND I MEAN OVERWHELMING rate of progress, advances, reporting, investing, and research going on in the AI field or probably ANYTHING that has AI on its name (can’t wait for the wave of mediocre “AI” startup scams that will be everywhere in the next couple of months once more and more people can mask an entire business as just an LLM with a couple of add-ons and seel it as a SAAS for unaware wealthy boomer executives(ok I will dial down the buzzword speak from here on out I promise) But what comes next will then news and business cycle and eventually the world be dominated by AI as the ONLY topic we will all be discussing from here on out in humanity as it apparently is starting to permeate in almost any part of human discourse?
Well I hope not, but also we might be biased because we are currently in the midst of what could be called the peak of the technological Hype cycle. But what will happen after that remains to be seen.
The Rollercoaster of Tech Hype
The Gartner tech hype cycle (or hype curve) is a graphical representation done by the tech and market research company Gartner ( a very creative name I know) that plots the stage that different emerging technologies have on a curve of how much Hype they have and in what stage they are. And this reflects a bit of how after a breakthrough some technologies become the hot topic everyone talks about, then it dies down and suddenly it becomes commonplace and it s no longer noteworthy.
Throughout our technological development, we’ve seen this; news, trends, and waves of investment hyping and elevating certain technological development over others. Sometimes the technology lives up to the hype, sometimes it goes above expectations, and other times it tanks horribly (remember the wave of 3D TVS and cinemas a couple of years ago?) And the cycle goes a little bit like this:
1.- Innovation trigger: This is when something is Brand new and people are still figuring out how to use it and trying to comprehend where we stand on it and how can it be applied in different industries. Is it good or bad, who knows barely anyone understands it yet.
2.- Peak of Inflated expectations: This is when things are HOT, and everyone nad their aunt is talking about it ( if this is a good or a bad thing who knows?) this is where we are right now with AI tools.
3.- Trough of disillusionment: This is when the tech doesn’t live up to the hype and the crypto bros are crying and losing all of their investor's money because Bitcoin wasn’t accepted by all central banks in the planet as its single digital overlord and we don’t radically revolutionize the entire global financial system in just 5 years. (so sorry bro)
4.- Slope of Enlightenment: Then after the depression more people continue developing the technology, maybe the regulatory framework changed, and finally a good match for the technology or a key input becomes cheap enough and we can finally grow and it comes back with a vengeance.
5.- Plateau of productivity: This is when the hype and reality are better aligned and the promises of technology are aligned with people's expectations because we’ve had time to understand its limits and its potential and aren’t in the honeymoon phase anymore.
If you want you could summarize this as when you start a relationship at the beginning you meet someone new you’re all excited and happy to see a text from them on your phone. Then you start going out and you start to put them on a pedestal of idealized and unreal expectations. Then you discover they are a flawed human just like you, you have your first couple of arguments, but you work it through then suddenly you are 2 years in living together have a pet dog, and are sending memes from your phone to each other sitting across the living room. (or something like that)
Thanks for reading The Placeholdr! Subscribe for free to receive new posts and support my work.
Why do we Hype Tech like this and why do we suck at forecasting?
I think on a more fundamental level it is part of human nature to get excited about things, people, technology, ideas, etc… let's be honest we’ve all been through this cycle in many different technologies. And because forecasting technological and economic development is hard, if not ney impossible to do it accurately and consistently (just ask any economist) we through around trends and hope for the best but most of the time the difference between a wild guess and an informed one is surprisingly low. (though it is better to have economists guessing than NOT I’d say)
And underestimating problems is not an uncommon thing even in the field of AI, damn the entire research and development of modern Artificial intelligence is birthed from a conference that happened in 1956 at Dartmouth University that was responsible for workshopping the ENTIRE development of artificial intelligence within 10 weeks because the organizer thought that was enough time to work and develop all of what could be necessary for that time frame.
This is crazy, but at the same time tells you how we tend to underestimate the challenges we might face ahead.
So right now we might be riding the height of the wave comfortably reaping the benefits from the current advances and developments happening at such a fast pace that timelines for advances in the field of AI cover mere months or even weeks where big leap forwards happen. But maybe in a couple more months or years, we might see the third AI Winter.
Winter is Coming?
AI winter is a period of reduced funding, research, interest, and overall Hype(god I’ve overused this word for this article) This happened mainly again due to misplaced expectations and the technology not being ready to meet up with the expectations that we had put into it. One of these winters came after a report in 1966 done by DARPA after they had funded for some years computer-based translation to be used during the cold war for military purposes. But this report published by ALPAC (one “A” short of a great name) concluded that computer-based translation was slower and more expensive so DARPA cut off most of its funding from AI in the 60s and it took almost a decade to defense contracts back into AI researchers.
This cycle has happened in some waves and some experts forecast the same thing could happen shortly, mainly because of some technical hurdle we don’t foresee at the moment, or after a while the market and benefits from this AI surge will be reaped and the development further ahead will be too slow or expensive to keep the trend alive with the same enthusiasm as we are currently seeing
Now will this be a good or a bad thing? It depends on who you ask, for some an AI winter would be a gift from the heavens because it represents a time that humanity can have to solve ethical and safety issues such as the alignment problem, as well as giving us enough time to create a more cohesive and logical regulatory framework for the development of a beneficial AI industry. For the other people on that spectrum, an AI winter would be a horrible destiny to be faced because it would mean a delay in solving the great problems that face humanity such as the climate crisis, poverty, and curing diseases such as cancer or aging.
Will AI be different?
Honestly, I don’t know a lot of people who claim AI is a radically different technology because we are creating an “alien” intelligence that can match and eventually outpace humans in its capacity for intellectual progress and technological development.
And others such as economists Philippe Aghion, Ben Jones, and Chad Jones (What a CHAD) propose a more level-headed forecast where it proposes different scenarios. One scenario is where AI just like most technological developments since the industrial revolution is a positive impact on economic growth ( a very large one) but eventually, the impact of a single technological advancement is good but in the big scheme of things it isn’t as world changing as we might expect, and even stretching that this could be a very revolutionary technology, again, after a while economic growth stabilizes and we reach a new plateau.
The other scenario proposed by Aghion, Jones & Jones is where AI leads to a great increase in economic growth but it keeps on accelerating and moving upwards and upwards without ever reaching a plateau. (similar to what David Roodman Mentions in his paper about a global economic explosion in 2047)
So are things gonna go faster or slower?
Hopefully, in a couple of years, the loud and chaotic world of AI news will stabilize and we will see a bit of a die down or at least a slowing down on reporting regarding AI. At least that's what I hope for the sake of everyone's mental health. I think we need something of an AI Winter or at least an AI Autumn to chill the fuck down for a minute and develop a robust ethical and regulatory framework for these technologies in the coming years.
I would like for us to overcome this HYPE cycle that is great that brings more people interested in a fascinating topic such as AI, but it can be detrimental when we have a wave of scammers and bad-faith players such as what we say in the crypto and web3 space that sadly has left a technology such as Blockchain that has a lot of potential with a stain such as the existence of crypto bros.
Because no doubt the exponential growth of AI is coming, let us not be mistaken about it. But that means that we also have a greater challenge and as researchers, commentators, consumers, and as a society as a whole we need to better understand and use these tools for our benefit and development. To solve humanity’s problems and try to create as little as possible of new problems. Also, our understanding and comprehension of the advances and changes AI will bring require more voices to be heard about the impacts that this technology will have as it pushes humanity forward into unforeseen territories that hopefully are filled with discoveries and rewards for humanity to continue exploring and growing as a species.