I recently had a wide ranging conversation with a colleague about the future of the internet and how the web will look 5 years from now.
If you look back even just 5 years, AI hadn’t yet gone mainstream. Go back 15 years, social media was just starting to make it’s mark on culture. In fact, every few years, there’s another new buzzword that a lot of the same people will jump on and position themselves as experts. There’s a hell-of-a-lot of overlap between the people who are loudly proclaiming that that the future belongs to Artificial Intelligence today, and the people who said that NFT’s were the future a few years ago…or Blockchain before that, or Augmented Reality, or Virtual Reality, or Chatbots, or Big Data, or the Internet of Things, or the Cloud etc. etc. ad nauseum.
Once you’ve been around for long enough, you start to recognise that, in most cases, whatever is currently being hyped as the next-big-thing, will be forgotten again before too long. For example, a lot of supposedly very smart people were convinced around the time of lockdown that Clubhouse (anyone?) was the future, of social media. As the ancient Greek saying goes, the only constant is change.
However, it’s often not too difficult to identify which trends will stick. Despite some people dismissing the early web as a fad, the world wide web always just felt like it was going to fundamentally change everything.
Conversely, almost everyone, except those who thought they could make a quick buck from it, immediately recognised that even the basic concept of NFT artwork was ridiculous (leading to the single most embarassing segment of television ever recorded)
All that said, the promise of AI does feel more like the former than the latter. There’s no doubt that, once the dust settles, the world will look very different, but the way AI is currently being shoehorned into every single app and service right now, in a desperate attempt to find a viable profit model definitely feels more like the latter.
A lot of tech companies appear to have lost their collective shit since ChatGPT first launched, in a desperate attempt to not be left behind.
Before ChatGPT, Google had, for years, been carefully and slowly rolling out some really smart AI augmentations to their services, giving the impression that they were significantly ahead of everyone else when it came to AI. I even wrote about this almost 10 years ago!
They claimed at the time that a slow rollout was the only responsible way of handling AI and had a publicly stated set of AI ethics principles. However, as soon as chatGPT came along, all caution was thrown to the wind, leading to their head of AI, the man dubbed the Godfather of AI, loudly criticising the company as he quit! Google immediately took the breaks off and launched their own AI chatbot called Bard, which was so disastrously under-cooked that they had to hit reset, give it a complete rebrand and relaunch it as Gemini less than a year later. (It hasn’t faired much better since then, inspiring an entire subReddit to poke fun at all the ways it regularly gets everything wrong ).
At first, it was an expensive add-on, but when they eventually realised that not nearly enough people were willing to pay, they just foisted the basic tier onto everyone (and increased the basic price of Google Workspace, of course), like it or not.
Ever since then, they’ve been using Gemini to ruin their own reputation by diluting search with frequently incorrect AI answers and prompting users, with increasing desperation, to consider using Gemini every single time they try and do anything in any Google property.
Microsoft
Microsoft had famously launched, then very quickly scrapped, the worlds first nazi AI (Yes, really!) a few years before chatGPT. (Depressingly, fascist AI’s do seem to be a thing). So when ChatGPT launched, rather than risk another PR disaster, they bought their way into the market by investing $1billion in OpenAI.
Soon after, they added ‘Copilot’ (Like some sort of cyberpunk Clippy) end-to-end on Windows 11 and evidently started a competition with Google over who could pester users with the most annoying popups, pleading with you to use their AI.
Apple
Apple, the company who watches other companies innovate, then launches their own take on an idea once it’s already mature technology, were clearly even further behind the curve than Microsoft. They launched their last iPhone with a flashy campaign showing how Apple Intelligence was going to run rings round all the other AI’s. At launch it was ‘coming soon’ and, nearly a full release cycle later, other than a few little gimmicky tools (image generators and the like) it’s still nowhere to be seen, so if you bought the iPhone 16 on the promise of Apple Intelligence, don’t worry, the iPhone 17 will be out soon and I’m sure that one will have it.
The simple fact is that, no matter how much utility LLMs may offer end users, no one has figured out how to turn profit with AI yet. All of these companies are hemorrhaging money and energy on the promise of what AI could be, beyond all conventional business logic.
They can only continue to operate by grace of investors, who have been convinced (by the tech companies, no less) that the returns are going to be huge as soon as they figure x out. (x can be whatever the investor believes that they should be doing!) There’s a very real chance that some of these companies are going to run out of runway before they ever figure it out. If you think that’s impossible, you are probably too young to remember the bursting of the dot-com bubble!
In Thinking Fast and Slow, Daniel Kahneman lays out how a combination of priming, the halo effect and confirmation bias can have the effect of convincing otherwise rational people to make a particular choice or action, even where evidence exists that a contrary action would be more beneficial.
This seems like as good a way as any of explaining how the entirety of silicon valley have been swept up in a collective fever dream when it comes to AI.
We’re talking about people who think of themselves as the smartest guy in any room they walk into, so obviously, they know what’s best.
They saw the ChatGPT launch; The most successful app launch ever; Zero to 100 million users in under two months. Those sorts of numbers lead to a thought process which is a combination of bravado…
“With all those users, there must be profit to be had, right? We’re geniuses, we can figure this stuff out.”
…and desperation…
“Everyone is saying that AI is the future. Our shareholders don’t want us to be left behind, so we’ve got to jump onboard too”.
So, AI is the future because every company is launching an AI product, and every company launches an AI product because AI is the future. This all leads to a perfect storm of circular logic, which completely overlooks the simple fact that the operating costs of AI are significantly larger than even the largest non-AI apps, so normal business logic does not apply.
Which leads to blind, baseless optimism:
“Hear me out here…What if…we let the AI get good enough to figure out how to reduce operating costs and make money?”
…And so it continues. 🤦
So, to return to the original question of what will the future of the web look like, five years from now. No one knows. Not the ‘geniuses’ of silicon valley, not the entrepreneurs, not the software engineers, researchers, bandwagon jumpers, hustlers or investors. No matter how much their bottom line depends on convincing you that they do.
I can state with confidence that I definitely have no clue. The only thing I know for sure is that I can’t wait to find out!
Photo by Steve Johnson on Unsplash
No AI was used in the writing of this article 😉
