Without a doubt, 2024 has been the year of AI. More specifically, it’s been the year of generative AI, a category that includes popular AI tools such as ChatGPT, Google Gemini, DALL-E, and Microsoft Copilot. As defined by Google Cloud, generative AI is “the use of AI to create new content, like text, images, music, audio, and videos.” You don’t have to be immersed in the tech world to see this trend playing out. Between the constant advertisements on television or the sea of articles highlighting its pros and cons, it has been impossible to escape the AI frenzy. While AI, including generative AI in some capacity, has been around for a while, it is only in the past couple of years that people have had such easy access to it. Not only is it easy to get your hands on, but it is also by far the quickest, most capable it has ever been. So it makes sense that countless companies, whether they are tech companies or not, have been engaged in an arms race to offer generative AI in their products and services. It seems like nearly everyone will tell you that AI is our inevitable future, but I’m here to evaluate that claim. However, I don’t seek to write an article on if AI is good for our society or not, but if it really can streamline our lives and open the many possibilities that the big tech companies claim.

A little backstory

If you could point to one piece of tech that was the catalyst for the recent AI boom, it is most definitely OpenAI’s ChatGPT. In fact, some people less versed in the AI world will often use the terms “ChatGPT” and “AI” interchangeably. In the early days, it was certainly more of a novelty than a useful tool. People would go and and enter prompts for fun, sometimes to see what the tech really could do or just to mess around and produce a funny result. The results would often be mixed: there were some things it was really good at, but there was also plenty that it struggled with. The writing style was often recognizable for its blandness, and queries outside of the AI’s capabilities were often met with the now iconic phrase, “As an AI language model…”. Still, the early results were promising, and the technology matured quickly. Soon, people realized this tool could be put to use for mundane tasks like writing emails, summarizing information, or to the dismay of teachers and professors everywhere, writing papers for students.

After ChatGPT became something of an overnight success, the other big players were soon to follow. On March 21st, 2023, Google released Bard (later to be renamed Gemini), an LLM meant to compete with ChatGPT. Instead of competing, Microsoft decided to pour money into OpenAI, investing $10 billion dollars in January of 2023. This led to the creation of Bing Chat, a tool built using GPT-4 integrated into the Microsoft Edge browser. Just like ChatGPT, these early models had many flaws (Bing Chat especially became completely unhinged), but once again the vision for an AI-powered future was becoming clearer by the day.

Fast forward to today, and AI is a much more mature industry. It seems like every company has found a way to integrate AI into their product. Grammarly has an “AI writing assistant,” Bank of America has an AI assistant named Erica, Snapchat has the conversational (and slightly creepy) “My AI,” and Notion has the creatively named “Notion AI” to assist your note taking. These all have varying degrees of usefulness, but in general this just feels tacked on. It’s like these companies added AI tools for the sake of adding AI tools, which allows them to use the AI buzzword in their advertisements and have something to please their investors with. In many cases, this integration only improves the user experience a bit or even not at all, but that doesn’t stop these companies. Generative AI, which once felt magical, is now just a part of our lives. This brings us back to the question that this article seeks to answer: will this really revolutionize our everyday life?

Powerful tool or overhyped novelty?

No one is doubting the huge potential that today’s generative AI has. The big models out there are multimodal, meaning they can process information of different types, such as text, images, video, audio, etc. It’s clear to see how capable and versatile this could be, but what are people really using this for in their workflows? Despite all of this potential, going out of your way to use AI is often just not practical or natural. For example, say you need to write an email to a coworker that you need to schedule a meeting at 3pm in the conference room to discuss the details of an upcoming project, and that the meeting should only take half an hour or so. It is this type of simple email that AI promises to be able to crank out for you so you can maximize your time, but think about if this is really saving you any time or effort. The email itself has to be just a few sentences, one that would probably take most people just a minute or two to write. Of course, AI will generate it instantly, but you need to think of a prompt that fits all the details in, and then you might want to edit the response just a little. Even once you’ve copied and pasted the response you have tweaked to your liking, it’s not too difficult for the recipient to see that you have sent this email using AI. While that’s not the worst thing in the world, it may not sit quite as well with the person who received the email that you haven’t even taken the time of day to reach out to them with your own writing. After all of this, how is it any easier than just writing the simple email yourself?

Here’s another example: you’re a student that needs to do some research for a paper you have to write on the Battle of Gettysburg. Asking Gemini or using the built-in Google Search AI overviews provides some excellent surface level information, but that’s about it. Of course, this isn’t any more useful than the first paragraph of the Wikipedia page. To get an A, you will need far more information, and Gemini prompts will only get you so far. Not only is it more difficult to acquire specific information this way, but you’ll basically have to double check everything it spits out for accuracy. After all, there is a disclaimer under the Gemini prompt window saying, “Gemini may display inaccurate info, including about people, so double-check its responses.” Also, you really can’t cite an AI model, so you’ll need a more traditional source of information like a textbook or website anyways. Once again, this has proven to be a little more hassle than the AI ads promised.

Speaking of ads, it should be noted that big tech companies seem a bit out of touch with the way people actually want to use AI, or perhaps what they think is normal and good. Anyone that had been watching the recent Paris Summer Olympics surely noticed the abundance of AI ads, especially from Google, who was heavily promoting Gemini. One of their ads in particular sparked controversy and resulted in the ad being pulled midway through the games. For those who haven’t seen it, the ad features a young girl who is a huge fan of Team USA track athlete Sydney McLaughlin-Levrone. The young girl wants to train to be just like her, so her dad uses Gemini to help him learn how to teach hurdle technique. This is a reasonable use of AI, so if the ad stopped right here, people wouldn’t be complaining. However, it is the next part that hasn’t sat well with everyone. The dad next asks Gemini to write a letter to Sydney on the behalf of his daughter telling her how big of a fan she is and how she intends to break her record someday. It should be easy to see the problem here: no athlete wants to receive AI-written letters from their fans. It completely defeats the purpose of fan mail, and the young girl featured in the commercial seems to have no part in this letter whatsoever. The ad is clueless and it is amazing that it made it past the focus groups and executives that must have had to clear it. While this seems more like a commentary on the ethics of AI rather than their utility, it comes back to the main points of this article, which is that these companies sometimes are out of touch with how people are actually using the tech they create.

Despite this critique, my bottom line here is not that AI isn’t useful. Clearly, there are use cases where it makes a lot of sense, I just don’t think it’s always in the traditional formats that might come to mind when someone thinks “AI,” such as chatbots or image generators. There have been big strides made recently in integrating AI more seamlessly into the way we already use technology, and despite how I blasted Google in the last paragraph for their cluelessness, they are definitely leading in this category. One such tool is Google Circle to Search, which allows you to circle anything on your screen (whether it’s text or images) and receive a Google Search (or generative AI) summary. This is an extension of Google Lens, which allows people to search using their camera, but it makes it even more convenient and capable. Another useful tool is Google Magic Eraser, a Google Photos feature that allows you to circle anything and have it magically erased from your photo. It works decently, and it’s useful for editing little imperfections or unwanted distractions out of your photos. It is AI features like these that feel like the future, and it is at its most useful when you don’t have to go out of your way to use it. Google is an excellent position to do this just because of the sheer number of services they offer (including Android), and this seamless integration could be one of the better use cases for AI in our daily lives.

Verdict

Now after arguments both for and against, what is my personal answer to the question posed in the title of this article? It’s definitely a bit more nuanced than a “yes” or a “no.” Some lofty claims have been made about AI, and I think that the era of the generative AI hype train won’t age the best in tech history. The romanticism for our AI future expressed by many in the tech world is just not realistic (nor is it the future we should really want). It won’t necessarily be the do-everything tool that it is often advertised as, however I believe that companies will get better and better at finding ways that people can meaningfully use AI once the novelty wears off. The future of AI will certainly be more than just chatbots. The technology is there, all that’s needed now is to find more ways that people can tap into that power to truly find utility in it. There is a long road ahead of AI, and just like any new technology, only time will tell what the future holds.