The technology world is going absolutely bonkers for AI at the moment. I get it—AI represents a technological innovation quite unlike any that I've seen before.
Admittedly, I don't know that much about it. I haven't thrown myself into it by any means, but I've certainly explored it a bit and I've taken the course that I usually do with emerging technologies... I've tried to consider the implications and potential of AI with reason and common sense.
The tech industry is maddeningly high on hype these days. A clear indication that I will always be something of an outsider in the tech industry is that I have always been nearly immune to the hype surround new technologies. I don't start giggling at the potential of what this could all mean. Instead, I've generally opted to let new technologies prove themselves to me. Show me something that materially improves my life or work? I'm all in. And I tend to remain all-in until something proves itself to be truly better.
Case in point? I'm still using Trello for project management 10 years after I started doing so.
I'm not saying that my approach is good or bad—I've certainly missed out on "being early" multiple times. But I will say this... of all the technological advances that have promised to be completely world changing in the 15+ years I've worked in tech, only one has truly lived up to the hype... mobile.
The fact that I'm even writing this post shows that I think AI has a lot more potential than other tech trends. But ultimately I think the value of this post comes from mindset—I'm looking at AI though very sober and objective eyes, which I don't see a lot of right now.
Here's a smattering of thoughts on what I see so far.
AI is both superlatively good and bad
I think that the most interesting part of this discussion is that AI will end up being both massively useful, and massively problematic.
- It will solve problems and complete tasks in a way that represents giant leaps forward in productivity.
- It will generate massive problems related to bias, plagiarism, spam, and the decline or our own cognitive abilities as humans.
Wrangling AI is going to be massive issue, but it will also lead to massive advancements that can't be dismissed. I don't know that I've ever seen a new technology with such truly polarizing upsides and downsides.
AI will dramatically accelerate software development
I work in the software industry—it's very clear to me that AI will significantly increase the speed at which software can be built. That's a huge positive.
That said, the tech industry hype is peddling the idea that great software will now be built in hours or days instead of years. This is not correct, and frankly the only people shilling this message are people who have never really built software themselves.
The young bucks can't contain their excitement—it's GAME OVER! Whereas almost every senior engineer I know is much more clearly thinking, "OK, where does this fit in?"
Writing code is only part of developing software—frankly, it's something that AI is well suited to. But it's only one piece of the puzzle. AI is an accelerant, not a replacement.
AI will make skilled humans more valuable, not less
This is probably the item on this list I'm most sure of. There's so much talk of AI taking peoples' jobs, and some of that will surely happen. But overall I think AI will make truly talented people that much more valuable across the workforce.
Stories that are deeply human and personal are what connect us. Everyone's creativity is unique to them. And even deeply technical topics need someone who truly understands them and has the experience to know how and when to apply them. These things simply cannot be outsourced.
A parallel that I love here the number of teenagers who learned to play the guitar really well in the 1970s versus the 2010s. Teenagers today suck, comparatively, at actually playing actual musical instruments. They've opted for Guitar Hero and electronic mixers instead.
The same could be said, generally, of the trades—plumbing, carpentry, etc. Your average 25 year old today has serious digital skills, yet can barely change a lightbulb.
No matter how smart or accomplished AI becomes, there's a market for people who truly understand the underlying concepts of whatever it is AI is trying to accomplish. Those people will be in shorter supply, which will make their skills that much more valuable.
A city of skyscrapers and no structural engineers
I think when most people consider their biggest fears or concerns about AI, it has something to do with bias. We're going to suddenly take whatever AI says at its word, when AI is really only as intelligent as the information it's been trained on. It will introduce bias as truth into our society in innumberable ways. While I think this is 100% true, I think this is happening already—AI or not.
My biggest concern about AI is the idea that we'll end up building a city of skyscrapers, yet have no structural engineers.
We will build stuff bigger, faster. AI will represent a leap forward in many fields, very quickly. But what happens when something goes wrong? What happens when something needs to be fixed—when an intervention needs to take place?
If you think AI will just solve all of the problems it creates, you've been brainwashed. We're going to see fewer people with real world skills and a deep understanding of topics like engineering because they've never been required to approach these subjects with the same level of mental rigor that previous generations did. There's a real potential that the things we build are a bit of a house of cards.
My dad—an engineer himself—witnessed a version of this throughout his career that I think was far less severe than what we'll see with AI. He started his career drafting plans for buildings with pen and paper, doing manual calculations for which materials should be used and how much weight buildings could support to be structurally sound.
Then came calculators, computers, Autocad. The more junior engineers he hired never applied the skills he learned in the same way—the answers were simply spit out of a machine for them. A huge advancement, right?
In most ways, yes—but it ultimately made my dad's skill set that much more valuable. His job morphed into often catching errors that the more juniors engineers missed—oftentimes errors that could have really catastrophic consequences.
The computer spit out a number that didn't make sense, that would result in a real vulnerability to the structure. But the more junior engineers simply didn't have the same level of experience or understanding of the underlying concepts to alert them that something looked a bit off. If the machine said it was true, it was.
While tools like calculators and computers definitely reduced instances of human error, they couldn't replace the human level judgement that came from experience and a deep understanding of the subject matter at hand.
There will be a massive recoil against AI
While I believe AI will become more integrated into our daily lives—and not just the lives of us in the tech industry—I think it's also fair to say that it will come with massive backlash.
Threads like this one demonstrate how AI can lead to massive productivity gains—an AI tool sent out thousands of automated, semi-intelligent emails on its own basically outsourcing the work that a sales agent might do. This is a great example of productivity gains!
But guess what? Doesn't everybody, almost unanimously, hate being on the receiving end of emails like this? When these emails can be employed en masse, that much faster, guess what's going to happen to our email inboxes?
I think this recoil will happen MUCH faster than most people think. We'll tire of the spam. We'll tire of reading AI generated articles. We'll tire of having shitty experiences with AI support agents and websites that clearly weren't built for a human.
I don't say this to be anti-AI... I say this because I think this represents almost as much opportunity—especially for start-ups—as AI itself!
They say when the rest of the world zigs, you should zag. I think a lot of people will become wealthy building things like AI detection tools. I think we'll see a huge increase in things like email products that only allow you to receive emails from people you've invited into your own personal email network.
People will crave the "human" element in an even more insatiable way.
I really believe that watching AI continue to unfold is going to be quite unlike most of the other tech trends that I've seen in my career. All things considered, I'm pretty bullish that it will lead to all sorts of positive advancements.
As for me, I won't be spending countless hours searching for applications for AI—but when obvious use cases emerge where AI is up to the task, I'll be jumping on those. In the meantime, there's a lot of "regular" work to be done that I know will deliver value to real people, in the real world.