Where is AI Headed in 2024?
A few prognostications that might be useful.
The more time you spend in the generative AI space the more you realize that there’s a lot going on but no one really knows where it’s all going to head. And, as much as we might want to believe that everything is going to change the reality is that humans usually over-estimate the impact of new technology, mostly because frontier tech excites us in a way where our imagination can easily take over.
I particularly like this quote that is oftentimes attributed to Bill Gates:
We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten. Don't let yourself be lulled into inaction.
The key takeaway here is that, regardless of your position and your understanding of how GenAI will personally impact you, your industry, and perhaps most importantly your professional work and future job prospects, is to simply not get “lulled” into inaction. Generative AI is here, it’s not going anywhere, and the impact is already being felt and distributed into your field of work — the question is whether you’ll be ahead of the curve or behind.
Still, it’s fun and interesting to capture prognostications from industry leaders and I’ve found one presentation that is somewhat long (87 slides) and I’ve done some work to simplify the messaging to things that are more practical in nature. Enjoy!
Ben Evans is a well-known venture capitalist who will create cogent and impactful industry-centric presentations that I’ve enjoyed over the many years that he’s been putting them together. I was just waiting for his AI one and he didn’t disapoint.
He actually uses a Bill Gates quote to kick things off too!
We all know that ChatGPT is OpenAI’s product and it is quite impressive at first blush — if you haven’t given it a go then it’s time to create an account and just fiddle with it and get an idea of what it’s like to “prompt”. The next slide is self-evident and obvious but why this is important (and the data-source specifically) is because the most exciting (modern) technology advancements come from the builders themselves and Hacker News is a key indicator of not just developer and engineering interest but a proxy for venture capital investment, spend, and job creation.
It’s going up and to the right and so, as I’ve said a few times already, follow the money (via engineering and builder interest and talent).
To the question of whether this technology is coming for your industry I think we all know that it is; the question is when. Here’s a snapshot that was a useful reminder:
And just like mobile and the web and cloud computing is everywhere the same paradigm is certainly going to play out. It isn’t hard to imagine that most of the tech products that you use (and apps on your phone) will have GenAI components either bolted-on or entirely built-in. In fact I’ve already seen some local
In fact, I’ve already seen some small attempts to bring GenAI right onto your own phone like “Offline Chat” which is a local bot on your iPhone:
Introducing Offline Chat, the next-generation AI ChatBot that runs entirely on your device without the Internet. You can use it anywhere, and your data stays private and secure.
The future is already here.
Some of the more well-known use cases have been captured by Ben below:
Early Wins: Code assistants, brainstorming, auto-suggest
Thin Wrappers: Add some UI and data linkage to OpenAI APIs
Synthesis: Point an LLM at your corpus and get answers
2024: What new problems can we discover?
The winners will be products and companies that solve real problems that people have, saving them time, money, and can be fine-tuned to be hyper-personalized, solving unique problems specific to each user.
But, like many cold-start challenges, the question is, well, where does one start? If the promise of GenAI is that it can do most (all?) things, then, are there going to be a million of these apps or a few dominant winners? Does it even matter?
Platform Shift: Actual product needs lots of single-purpose tooling, UX, and TGM, just as before.
General Purpose: Software that can do “anything”
Paradigm Shift: Far more tasks and jobs can be automated by far fewer apps
Enterprising entrepreneurs will find solutions to many of the existing problems that each industry has but there will be a lot of duplication amongst the players and if capitalism prevails then there can only be a few at the top who have “monopoly” power and the revenue to sustain into a generational company.
The rest of the presentation focuses on AGI which isn’t as practical for this newsletter but feel free to peruse that section if you’d like. I’d like to end with one of my favorite quotes about technology via William Gibson, a science fiction writer who pioneered the subgenre and who coined the turn “cyberspace” (and continued to “cyberpunk”):
The future is already here — it’s just not evenly distributed.
I think that applies to generative AI almost perfectly.
Oh! And finally I was sent some images via one of my readers who successfully installed Stable Diffusion XL on their new Macbook Air while sitting in their Tesla — I love to see others find these posts useful! Here are some pics:
Looks like it’s good to go! Congrats! Here are some of their first outputs: