Why AI Is Tech's Latest Hoax

Right now, Kajabi is offering a free 30-day trial to start your business if you go to kajabi.com/modernmba 0:00 Modern Silicon Valley 6:42 Sponsor Break (Kajabi) 8:29 Post Dot-Com Beginnings 17:22 Data-Driven FOMO 28:19 Shovel Economics

As with any gold rush, the only one making money is the one selling shovels and beer, which in this case is Nvidia.

Starbucks is confused about its precipitous loss of revenue in 2024? I'm not. I'll tell you EXACTLY why. Everything that was unique about the Starbucks experience - ordering a cup of coffee, perhaps a bakery item or a lunch (love their PB&J meal) sitting down on one of their comfy couches, propping my feet up and doing work on my laptop, or just browsing the Internet, or even just meeting a friend for coffee, all disappeared during the pandemic. They took out all the comfortable seating, removed the electrical plugs for your laptop, and replaced everything with low, long wooden tables, hard wooden stools with no backs to sit on, reduced the seating to almost nothing, and basically turned Starbucks into just another place to get an overpriced cup of coffee. The business model is now "Order your coffee by mobile app, come pick up your coffee, and get the f*ck out. Next in line, please!" Hell, I can just brew my coffee at home cheaper and have a more pleasant experience.

As a computer programmer that has since the 1980's seen dozens of fads come and go I completely agree. We programmers will continue to do all the work and the talking heads will chatter about whatever their latest misunderstanding about technology is. Some fads are incorporated into day-to-day programming work but most just fade away into oblivion.

I was not expecting this video to strike such a chord with me. I've worked as a software engineer for over a decade, and I had this overwhelming feeling my job was complete bullshit for most of that time. We use bad technologies simply because they're popular, and most of the engineers don't really understand how computers actually work. The section about people being extremely tribal and only learning things because it pads a resume is 100% the truth. It's just a disgusting industry to be a part of if you have any integrity, just lies lies and more lies. I hadn't heard a great explanation of why it sucks to work in tech until I came across this video. Thank you, from the bottom of my heart, for making this.

I 3d printed an direct to consumer self driving all electric ai block chain app, I'm worth 80 trillion dollars now.

This video made me realize something for the first time. Every company is trying to use big data and/or AI to "understand the consumer", with the goal of trying to sell their crappy, overpriced product. So instead of focusing efforts trying to make a good product, for a reasonable price, they are trying to sell me their crappy product for as much as possible, and hope that AI tells them how to do that. I think customers are easy to understand - just don't rip me off man. No wonder most of them are failing.

Those corporate talk videos made me feel physically uncomfortable.

The funny part is that even appliances like washing machines, microwaves, etc have become 'AI Enabled' 😂😂

I am a Data Scientist. I would like to chime in on this fascinating topic. I was a Software Engineer before I transitioned to Data Science. To me, the big scam is not AI it's BIG DATA. Does anyone know that through statistics, the difference between a sample and a population determines whether you think big data must be used or not? The reality is big data is NOT required, and let me say this again, big data is NOT required to generate predictive analytics. What is needed is TARGETED data and that is a big difference. Now you do in fact need some data but I can turn out a very compelling machine learning model using 100,000 records of data if I can get a sample size that represents the population. That is the key to reducing the overfitting of a ML model which is one of the most challenging aspects of developing a successful prediction model. You need to know the difference between a statistically significant SAMPLE which is the key to machine learning. So, if you don't know any better, you believe that you must spend thousands of dollars on some nonsensical cloud solution to host 100 million records of data for analysis. That is utter BS. Give me a distribution with a standard error of .01 to .10 and I will give you a VERY accurate ML model with less than 100,000 records.

Once I started seeing generic airport ads from big dinosaur companies mentioning AI I knew it was dead

I was told by a CTO that my proposal to cut IT department expenditure by over 50% at our company by moving away from those Tribal providers, was DOA b/c when this company failed everyone had to get another job and having those ‘industry standard’ solutions on their resume was the only way to do it. Company went bankrupt about 18 months later. But to his credit we all did get other jobs.

The funniest part of all of these hype cycles is watching people jump from one hyped up product to another. Crypto, Ai, drop shipping, ChatGPT, and the rest. I like how they always insult others for just being rational about the technology being useless or, at the very least, not good enough as it is. They just forget the broken promises of the last piece of tech and jump to the next one.

After 20+ years in IT the only consistently performing tech innovation I have seen is Bullshido.

Rory Sutherland makes an interesting point about this: despite all the data being collected about you from every website, personalised advertising has largely failed. Online ads are 99% irrelevant to me. They can predict my demographics and some topics of interest. But heck, even TV and magazine ads could do that. There is none of the hyper-personalisation that's been prophesied for 30 years, that the machines would understand my preferences better than I do. Same with content recommendations from Facebook, Spotify etc. They also cannot predict any new things I might be interested in based on my personality. They just just shovel more of what I've once interacted with. That reminds me of another relevant Rory quote: "All big data comes from the past."

Anyone else notice the janky editing in the middle-ish third of the video?

The idea of more data = better solution is so flawed from the start. If you have more garbage, you would still get more garbage, not better solutions.

Corporations are so eager to be first to market with things like AI that they don’t spend enough time upfront to understand how it can actually fit in and drive value for their brand. Just release the AI deck and we’ll figure it out along the way. It’s bad.

As a former data scientist its funny to hear so many companies talk about how they thought thier data would be a game changer. Most companies didn't realize the limitations of thier in house data sets and didn't realize that they were collecting useless info. They didn't realize how they needed to collaborate with competitors and other companies in order to have robust data that would actually be useful. But even in doing all that I think the inherent issue with relying on data is that it looks at the past to predict the future which will always be faulty

I worked in tech venture capital for 10 years. What I saw was a whole crop of “world changing” concepts that one by one, failed to change anything. But at the end of that ten years we were talking about a whole new idea of what was going to change the world next. I realized all that talk never got us anywhere.