AI Won’t Make You Lose Your Job, Yet
This past week has been a busy one for me. In addition to us finalizing and working out of our office space, we’ve had a few tech issues come up that we’ve needed to fix, and I’ve also given two different talks, one on AI and one on startups in general.
Getting out there and speaking with the community has been an enlightening experience for myself. There have definitely been some common trends. Inevitably, whenever I talk about what I do, someone asks a question about Bitcoin or the blockchain. If you’ve seen any of my other posts, you know I have a lot to say about those. In addition to questions about blockchain, I get a lot of questions about data collection and cleaning, where AI is going, and also whether AI is going to take everyone’s job away.
It’s no secret that automation is affecting virtually every industry right now. From automated kiosks to self-driving cars to front office reallocations, automated processes are making business operations ever more efficient. Given the noticeable progression of AI and automation, a lot of people have been asking me what I think AI is going to do to jobs.
This is obviously a loaded question.
If you take a look back throughout humanity’s history, a common thread across all cultures and time periods is the development of new innovations. It’s true that many of these new innovations made prior jobs obsolete. But usually, what I’ve seen, at least, is that this obsolescence happens in slow waves, rather than in quick spikes.
Various countries and companies are faster to catch on than others. Older workers retire, and companies may choose to replace their productivity with machines, rather than new workers. Automated methodologies may work better in some areas than in others. In virtually all areas, what you tend to see is automation improving quality of life while slowly phasing out obsolete tasks.
I believe that AI will be no different, at least when considering the next 20 years, after we get to the singularity, all bets are off.
But if you just look at the foreseeable future, I bet you’ll start to see a lot of the same patterns that you saw when other major innovations became mainstream.
Large, existing corporations will likely take longer to change their processes, allowing workers from older generations to retire. Even if they can automate much of what humans do today, corporations have a lot of legal, regulatory, financial, and HR considerations to take into account when making far reaching changes.
Smaller startups will likely start to use many new AI technologies first, but those very same companies would likely never have been started with AI available to propel them forward in the first place.
Even if corporations wanted to replace workers en masse, they likely wouldn’t be able to, not given the state of AI today.
Even though AI has come a long way, we’re still far from the days where AI can make a simple judgment call on a problem that has multifaceted issues. A common theme I’m seeing among the people I talk to is that there’s this feeling that AI will be better, faster, stronger, and more importantly, smarter than humans in short order.
Having worked in this field for a while, I don’t believe that’s going to be the case. The AI that we’re developing today is optimized for very specific tasks, and even though they can do things that only humans could do until recently, they’re a long way from being able to “think” and solve complicated problems.
I think they’ll get there, but I think it’ll take a while, perhaps longer than any of our existing lifespans will allow us to see.
There’s no question that AI, driven by machine learning, has made gigantic strides in a short amount of time. But those strides come in spurts, with deep learning, reinforcement learning, and transfer learning having just now stepped into the spotlight.
And even the brightest innovators in this space acknowledge the massive shortcomings in the underlying methodologies of these learning algorithms. So for now, I think the best thing to do is continue to learn new skills and improve yourself, be aware of how you can use new data-driven processes to do your job better, and keep on keeping on, that is, at least until Google comes up with a machine that can tell you otherwise.