AI anxiety – 7 reasons for calm amidst the bullshit

It’s everywhere: “It’s going to revolutionize everything!” “It will replace human workers! “It will propel humanity into a new age unlike anything we’ve ever seen!” “If you don’t get on the train you will miss out!” And on and on.

It’s all (mostly) complete bullshit.

Before I continue, a few small but significant notes of my background.

I have 2 bachelor’s degrees, both in engineering. Before becoming a life coach I spent a decade in software engineering and led production teams responsible for worldwide mission critical projects. My last role before leaving the industry was as a Senior Software Engineer. If you’re curious about specifics, check out my LinkedIn profile.

In the past couple of years I’ve spent a (probably ridiculously) large amount of time researching the AI craze for reasons like out of pure curiosity, gathering investment information, trepidation and anxiety about the future, general interest in tech, and general interest about the world, people, and societies.

With that out of the way, I’m going to explain why AI need not induce fear/anxiety/angst/etc. Unlike most of my content, this is going to have a very annoyed bent to it. Because what’s going on is really really annoying.

Let’s start by understanding how Ai works in laymen’s terms:

A shitload of infofmation is fed into a bunch of servers. By “shitload” I mean “most if not all of the world’s knowledge”. And if I may add, this is often done by bulldozing and not giving a fuck about copy-writes or intellectual property.

Next, that information is organized and relationships are created. For example, a “car” is also a “vehicle” and a “motor vehicle” and a “mode of transportation” and a thing that needs “gasoline” but sometimes it uses “electricity” as well. It has *4* “tires” and *1* “steering wheel” and so on and so forth.

So, what’s been done here is a bunch of relationships are formed and attributes are created. Often, humans are used to help with this. BTW, you know the reCaptcha “identify all the motorcycles in this picture” prompts to prove you’re human? Those are used to train AI in image recognition. Congrats, you’re doing free labor for big tech.

So, what happens when you ask an AI “how to change a flat tire?” The AI runs through its databases and retrieves all information it has on the relationships having to do with “tires”, “flat tires” and “changing” said tires and now it does some probabilistic computing. Probabilistic computing just means that it guesses what the most likely answer should look like based on similar examples it has in its database. What word should come first, then what word should come second, and so on until it’s put together a response.

That’s it. The end. There is no “thinking” or “intelligence”. It’s just guessing shit. It’s not bad at guessing the correct thing but it’s also not great. Don’t believe me? Ask an AI about something you’re an expert at and see what it spits out.

Now that we understand how it works, let’s jump directly into the 7 reasons for calm:

Since AI is probabilistic, it makes mistakes. In tech lingo they’re called “hallucinations”. These fuck-ups are presented with a high degree of authority making them sound like all is good. But it’s not. I have a bunch of real world examples below for you to enjoy.
Since AI is nothing but probabilistic relationships, it means there is no *thinking* involved. It’s not “smart”. It doesn’t actually *know* anything. You’re way better.
A recent MIT research study (which can be found here) showed that 95% of corporate AI pilot programs have yielded no meaningful improvements or efficiencies or monetary benefits. Uh-huh.
AI is here to stay. There *are* a few areas where they’re effective and people will figure out how to make them useful in more applications. But this isn’t an overnight change. When computers were first introduced into the workplace in the 80’s, they didn’t change things dramatically. In fact, it took about 2 decades for companies to start seeing productivity gains. So, AI is here. But, let’s all take a deep breath – things are not going to be much different tomorrow or next year or in 5 years.
There is no actual intelligence. There are many definitions of intelligence but the one I use and consider the gold standard is *original thought*. Novel ideas. AI is incapable of anything like this.
Let me give you an example:
- If you ask a chef to create a completely original hamburger, the chef will use all his/hers cooking experience to think through which ingredients not normally used in burgers could make a good addition (or substitution for something else). They will take into account how different food ingredients make us feel – a hearty meal is vastly different from a sweet treat – and decide what would make sense for a burger. What ingredients are currently available? What sort of change to the classic burger recipe most makes sense for the restaurant and its patrons? They may consider presentation – how would people react to (for example) a purple hamburger? Can a purple burger become a hit with the right marketing? I’m not a cook so there are likely a plethora of other considerations I’m missing but you get the point.
- Now let’s look at AI’s answer: it sifts through all the info it has about a burger and since you asked it for something original it will then sift through all the info it has on what makes foods original. And spit out some random ingredient based on nothing but probability. Will the result taste good? Would anyone eat it? To an AI its immaterial because it has no fucking clue about anything. It’s a very powerful word processor and that’s it. The result may be amazing or may be garbage.
All those commercials and demos of “agents”? Yeah they’re bullshit. What they don’t tell you is that what’s happening is: a human is prompting an AI until it gives a response the human likes. Then the human takes that response and does something with it. And when that part of the task is done, the human goes back to the AI and prompts it again until the human receives a response they’re happy with and then the human goes and does something with that response. Rinse and repeat. There aren’t any god damn agents running around doing cool shit. That “agent” that hired a Taskrabbit? Yeah that was bullshit. But even if we blindly trust that these “agents” are doing miraculous things on their own, they’re not actually good at any of them. Here’s a Salesforce study about their effectiveness.
So what you’re saying is that AI is a whole bunch of hype? Yes that’s exactly what I’m saying. Like Blockchain, Crypto, the Metaverse, etc. It’s a cool technology that’s useful in some areas but because tech no longer innovates, it’s hyping it up as the next big thing so they can appear relevant and mint new billionaires. And because most CEOs have no fucking clue how it works, they roll with the “it’s the next big thing” narrative for fear of missing out. And because most media have no fucking clue how it works, they regurgitate the bullshit that Silicon Valley is spewing. There’s no intelligence going on anywhere including in the heads of CEOs and the media.
Bonus! All the talk is about AI getting better with more scale. More info. More training. More processing, more power, more money, more, more, more! That’s a lie. There are already signs that AI is reaching the peak of its abilities. The only “more” there is is money for the conmen and destruction of environment for the sake of city-sized data centers. There are many articles and data and studies that confirm there’s no more scaling to be had even you cover the entire Earth with a giant goddamn data center. Here is just one of them.
Bonus #2! Since everyone is using AI and AI will mostly spit out the same thing to everyone, being YOU, being your AUTHENTIC human self is an advantage. YOU are original. And in a land of the same old same old, original wins.

The point of all this is that yes AI is here and it’s here to stay and it will change things. But, let’s be rational. It will not spawn a new golden age for the few very rich while leaving the rest of us to fight between ourselves for a meager existence. Email was revolutionary. But we’re all still here. Think about that.

For an extensive deep dive into all this, I suggest you read Ed Zitron’s (tech journalist) lates blog for The Case Against Generative AI.

Intelligence?

Now for the fun stuff. A couple of examples of AI “dominance”. The first is what happened during a recent roadtrip to the DC area. Google had just made a bunch of updates to use its Gemini AI on Google maps. Here are some gems I took screenshots of while driving straight on a single Virginia highway:

The second example comes from me trying to figure out the best time for posting content on several social networks. I asked ChatGPT, Google Gemini and Perplexity (randomly labeled X, Y, and Z in the images below) the same question: give me the top 3 days to post on social media (LinkedIn, Facebook, Instagram, Threads) and rate them from best to worst. Also, give me the best times for each day.

I then logged into each of those networks and looked at my analytics data to see when my followers are active and the times of day during which they are most likely to be active. I then ranked those from best to worst and added the best time of day – this is the blue column. The results are below. As you can see, there is no consensus (if AI is so powerful and all-knowing, all 3 should have given the same answers) AND, none of the answers match the reality (my analytics from the various platforms).

Despite what you may have heard, we lowly human worker bees are still the top dog and will be for a long time. So rest easy, you’re not being replaced. This isn’t the Terminator and there is no Cyberdyne Systems corporation. And it’s not the Matrix either. Don’t believe the hype – it’s just there so a bunch of assholes in Silicon Valley can become insanely rich. There is no intelligence – it’s just algorithms that order information. It’s ok to be calm. Educate yourself and stay calm. Things will be ok.

Oh yeah, one more thing. The entire AI industry is not an industry. It’s a handful of companies passing the same money amongst each other and claiming record profits. Here is graphic of the relationships between the AI players. This is not what a healthy industry looks like – it’s incest. Credit to Anthony Restaino.