AI the next technology cycle?
I recently read Hardcore Software from Steven Sinosfky. The book explains in details the rise and fall of the PC era and, more precisely, the technology cycles that enabled it and the ones that caused its downfall.
The story is now known: the PC market boomed in the late 1980s/early 1990s and everybody wanted a computer at home. Microsoft saw an opportunity, executed (almost) flawlessly and shipped an operating system, a productivity suite and more globally, a consistent ecosystem. Money was flowing and users were locked in the Microsoft ecosystem.
What caused its downfall is two combined cycles:
the internet era: the industry moved from native/desktop to web-based applications. Starting in 2000 and onwards, a Windows computer was no longer required: all you needed was a browser. The entire Microsoft ecosystem (developer tools — such as Visual Basic — to the productivity suite — e.g., Office) was under attack.
the mobile era: with the iPhone, no one requires a computer at home — everybody has one in their pocket. It was impossible for Microsoft to turn the ship around and ship a good mobile product:
the Wintel alliance made them stuck with Intel — the wrong partnership for mobile
Windows has constraints that made it difficult to make a good mobile experience (UI impossible to translate on mobile, impossible backward compability. etc).
Microsoft was so blindsided by their arrogance that when they had a mobile product, they thought they would kill the iPhone and held a funeral after launching the Windows phone.
What Microsoft really needed is a reset of its product, start from scratch and design a mobile-first product from first principles. Doing so is almost impossible in a large company unless you have strong, product-focused leadership (like Jobs who shelved all the Macintosh lines when he came back). This ultimately came from Nadella who turned the ship around and focus on the cloud business.
Technology cycles come and go periodically. They re-shuffle the cards in the industry and wipe out entire products and companies:
mobile phone killed the camera industry (remember Kodak?)
internet killed many brick-and-mortar stores (remember Blockbuster?)
New products (e.g., iPhone) replaced the old one (e.g., cameras) because they provide a better value. Netflix is way superior to Blockbuster in any way: you can watch whatever movie from your couch at a lower price without having to pay penalties for returning the movie too late.
However, I often wonder if AI is the next technology cycle.
While there is a lot of excitement around everything AI-related, I struggle to see a significant value from AI products. I see the value from a subset of products (e.g., chatbots and ChatGPT like). It’s still hard for me to see any value in 90% of AI-based tools.
Is AI an incremental improvement to an existing product (e.g., Perplexity improving search engine) or a disruptive technology (e.g., something totally new like ChatGPT)?
There are some good products (e.g., ChatGPT being the biggest one) and yet, they need to prove they can become profitable (e.g., OpenAI is bleeding money).
And for each good product, there are many bad AI-based products that fail to delivery significant value (e.g., rabbit or humane being the most famous).
I am being told every single week to try a new AI-based developer tool. And the result is the same every single week: an over-hyped product without much value.
Today, it seems that AI is in the same bucket as blockchain technologies in 2015~2022: an interesting concept that needs to mature and prove value. Crypto/blockchain failed to prove value over the last decade. I wonder if AI will be the breakthrough everybody promises, or an incremental improvement to integrate in existing products.