AI: All The Hype And None Of The Results

Welcome to this week’s Deep-Fried Dive with Fry Guy! In these long-form articles, Fry Guy conducts in-depth analyses of cutting-edge artificial intelligence (AI) developments and developers. Today, Fry Guy dives into how AI models have failed to meet the excessive hype. We hope you enjoy!

*Notice: We do not receive any monetary compensation from the people and projects we feature in the Sunday Deep-Fried Dives with Fry Guy. We explore these projects and developers solely to showcase interesting and cutting-edge AI developments and uses.*

Is artificial intelligence the most over-hyped technology in history?

Over the past couple of years, we’ve been inundated with stories about how AI is going to be the next Industrial Revolution, leading to all kinds of prosperity.

People have said that AI will allow you to retire, and you’ll never worry about money ever again … AI will make you live forever … AI is going to cure every disease on the planet … and AI will bring back the dead!

According to the AI hype, you should book a cruise, have your robot butler mix you a cocktail, and tell your boss where to shove it. The government will just send you checks, right? I mean that’s what all the YouTube personalities are telling us. According to media outlets of all kinds, we will soon be able waste our lives away scrolling TikTok and playing Wordle while AI does all of the real work for us.

But is this actually going to happen? Will AI make everyone’s age-old worries—like paying the bills—vanish in the next few years? Should you start preparing to quit your job and travel the world for free? Or is the AI frenzy the most overrated breakthrough of all time? Let’s explore.

Since 2023, there has only been one thing more nauseating than the news coverage of Taylor Swift dating Travis Kelce. That is, the incessant narrative of how AI is a going to change society as we know it. In just a little over two years, AI sensationalism has increased over 1000%. In fact, AI is so overused as a word that it was named the Collins Dictionary Word Of The Year for 2023 (is “AI” even a word?). Somehow, everything even remotely associated with computer functions has been dubbed, “AI.”

But this AI fervor compared to the reality of what AI can actually do right now are worlds apart—it’s a night and day difference. Since the launch of ChatGPT two years ago, AI has continually overpromised and underdelivered. This is prevalent all over the AI landscape. OpenAI and Sam Altman, for instance, just released their new model called OpenAI o1, and it’s been touted as “insane,” “game-changing,” and “groundbreaking”—these are the usual annoying hype-words associated with new AI releases. One would think that if all these developments were so groundbreaking, there would be a rise in earthquake frequency, after all.

The problem with this new o1 model—and other new models like it—is that for the everyday user, there’s hardly any difference between using this model and older models of ChatGPT. Actually, a lot of people think ChatGPT’s latest o1 model is worse than its predecessor, GPT-4o. Try it for yourself! This “insane” LLM is sloooooow—it sometimes takes up to 50 seconds to answer simple prompts. Plus, the o1 model’s outputs seem poorer than GPT-4o’s in a lot of cases. In fact, it doesn’t even seem that much better than GPT-3, which was released two years ago. But the news outlets told us AI was supposed to be 100x better compared to 2022. So what happened?

Before you start freaking out with objections and say, “You idiot—ChatGPT’s newest model is on the top of the AI benchmark leaderboard,” let’s slow down a little bit. Almost no one—except for some know-it-all Reddit nerds—care about benchmarks (sorry, know-it-all Reddit nerds). What most people care about is how new AI models perform in the real world—how they measure up against the hype and help solve real-world problems and make life easier.

To be clear, the point is not that AI models failing to meet expectations is an issue. The technology is insanely advanced, and it’s incredible that we have the capabilities we do. I mean, in just one click, AI can create original images and write entire books! The problem, rather, is that the hype is overblown. The fact that AI will cure every disease, lead to universal basic income (UBI), and put everyone out of a job is (so far) one of the greatest scams in modern day history. But the weird thing is that nobody seems to care … maybe it has something to do with peoples’ investment in NVIDIA—their wallets are incentivized to go along with the crowd.

You might be thinking, “Okay, OpenAI’s o1 has failed to deliver. But other companies and AI models are killing it!” But this just isn’t the case. It’s not just OpenAI’s o1 that has failed to deliver on the hype. Google’s Gemini model has produced racist content in the past and had to be shut down for a few weeks. Google’s AI Overviews was supposed to overthrow traditional methods of searching, but then it told people to eat rocks and jump off bridges to solve depression. And what about the elephant in the room, Alexa and Siri? They are a complete laughing stock, often misunderstanding queries and sounding more robotic than ever. Those ladies haven’t improved since their releases. Apple Maps? Don’t even get me started—it can barely detect a traffic pattern or provide an accurate ETA.

Despite these big tech companies underdelivering on AI products, they still continue to promise big. Apple has promised AI updates of all kinds, but the company continues to push them back. In fact, the iPhone 16, heralded as “the first AI iPhone,” barely has any of the features it was originally touted as having. But hey, it has a “new and improved” camera and color scheme.

OpenAI’s Sora text-to-video tool was ushered out a red carpet and said to mark the end of videos and movies as we know them. Yet, months and months later, it’s still private and unavailable to the public. And remember GPTSearch—the engine that was supposed to “replace Google”? Newsflash: it didn’t. Instead, the project was a complete bust and got brushed under the rug so nobody would recognize its massive failure to meet expectations.

And what about AI taking jobs? Forbes said AI will take up to 85 million jobs by 2030. Elon Musk predicted that jobs will eventually become “kinda like a hobby” and that “the AI and robots will provide any goods and services you want.” Despite these predictions, only 5% of jobs lost so far this year were lost due to AI. Much of this is because AI is not (yet) able to do tasks at the same level of humans. So even though rumors about AI taking jobs have been spreading like wildfire, most people are still working the 9-5 and beyond. Now, that’s not to say many jobs will not be replaced by AI eventually (if you wait long enough, surely this will most likely happen), but it might be worth tampering expectations on timelines until there are more substantial results.

As you can see, the “all-powerful AI” that we’ve all been promised for the last two years isn’t even close to meeting expectations. And because new AI tools are underwhelming, large companies are having problems selling them both internally and externally. Businesses throughout the U.S. are canceling their Microsoft Co-Pilot subscriptions. ChatGPT’s growth has also slowed exponentially—their current user base is only a small fraction of Facebook’s or TikTok’s comparatively. And despite what commercials tell you, nobody (I mean nobody) is buying Apple’s Vision Pro. Apple expected to sell 800,000 units in 2024, but in reality, they are yet to sell 100,000. Many news outlets are calling sales of the expensive product “all but dead.”

So that brings us to our current situation today. Instead of AI taking over our jobs and letting us quit so we can gather UBI checks from the government, we’re stuck with some glorified AI chatbots and image generators—ones that can spell check for us and create poor blog posts.

At this point, AI has turned out to be a collection of basic play toys that distract us throughout the day. They can help us create some funny memes and get a C+ on our college essays, but this is a far cry from the supposed next “Industrial Revolution” that we’ve been promised.

So regardless of how many promises you continue to be told by tech leaders and news outlets, don’t plan on quitting your job any time soon. Instead, do whatever it takes to butter up your boss for a raise. After all, inflation is pretty crazy these days.