Unless you've been hiding under a rock lately, you've probably been overwhelmed by the deluge of AI news out there, good and bad. Unfortunately, it's very hard to tell truth from fiction. Much of what's out there is carefully controlled hype from a combination of good AI marketers and indiscriminate journalists. We aren't going to go into every nuanced area here, but we want to bring out the truth in a few key areas.

First, will AI take all of our jobs?

No. Never in human history has a technological or cultural shift caused expansive and long-term unemployment. However, changes like this typically do cause shifts in the work environment. Folks who know how to work with the new technology typically do better than those who don't. It may cause some shifts in career trajectories for some, but overall most believe that AI will enhance our abilities - in a "standing on the shoulders of giants" way.

While we are currently seeing a slowing of hiring in junior technical roles, that will turn back upward eventually. American businesses are usually very short-sighted. They look out 6-months at most. Right now, they are battling so much economic uncertainty that they are trying to control whatever they can. Junior employees are still learners and typically don't add as much value as senior employees. So if you can curtail some of that spending on juniors in favor of seniors, you can make the numbers work better - for the short term. We don't get seniors without having juniors, though, so companies will have to turn that hiring spigot back on at some point. Those who don't will be in a world of hurt 3-5 years from now.

Second, is AI actually thinking?

Again, No. AI is a predictive function and that chatbot or LLM you think you are talking with, is just using a large collected sum of prior work to predict what to say next. There is no understanding happening inside your computer. This is why we see errors in AI generated images and videos and mistakes in content returned when we ask a question (i.e., "hallucinations"). The computer is just trying to guess the next word to say based on how it was trained. It doesn't comprehend what it is saying.

Third, should I even pay attention to AI since the bubble will eventually burst?

Yes, you should. If you are old enough to remember the dot com bubble bursting, you also know that the internet didn't go away when it did. ECommerce didn't cease with the bursting of that bubble. Things look differently now, though. The internet of today isn't what we expected if you asked us in the mid-1990s,

With that, the people who paid attention and learned how to work with internet technologies are the ones who were successful, and some wildly successful, in the mid-2000s and forward when the public started pouring onto the internet en masse. People who shunned the connected technology were generally put out to pasture and pushed from the workforce, although that took many more years.

If we do see whole job roles being made redundant by AI, it will take a long time to work it's way through our economy. Blacksmiths didn't just disappear overnight. They actually shifted to a much more specialized role. We watched the roles of secretaries being gradually phased out over decades. We'll probably see the same with AI, if there are roles that are no longer necessary.

Finally, if AI can do everything a person can do. What hope is there for us?

Well, AI can't do everything a person can do - so take a deep breath and relax. As we stated above, AI is just predicting output based on previously collected works (i.e., its training). It's not creating new, novel, or unique works. The AI tools available today are really just a next step in automation capability. You should learn to use them that way.

They are much easier to instruct on tasks than having to write code. You don't necessarily need to have all your data nicely structured. It can figure it out. You don't always have to be complete in your instructions either. It can figure out some of your meaning from it's training and prompt you where it can't.

What you don't always get, though, is the same response to the same question time after time. You can learn to make AI more deterministic, but it's non-deterministic by nature. Math and guessing will do that.

Here's our suggestion: dip your toes into an AI platform and learn how it works. Learn the concepts and capabilities. Learn how you could use it in your workflow to get more done (that's what we are doing). But... learn how to do it securely.

With all the hype comes a lot of "trust me, bro" from the hypesters. Unfortunately, they aren't the ones who are going to hurt when your AI goes off the rails. Check it's work, give it minimal access, be careful in select vendors when you start testing agentic use. Be careful about the types of information you plug into that "chat" window.

While we will have it generate code all day long and reason through business functions, we don't trust it with any sensitive information. And always be a human firewall between the AI and your business or personal life. Don't just let it run amok with your business or life.

In short, baby steps into the cold pool water. No cannonballs.