Searching for Answers?

So, you may have heard that artificial intelligence (AI) is, well, happening. What you may not realize is just how much it’s happening. McKinsey & Company conducted a market survey, which revealed that self-reported AI utilization for one or more functions has increased from 56% of respondents in 2021 to 78% of respondents in late 2024. That doesn’t include any data from the current year (which is nearly over, yikes). Stanford reports a whopping $109 billion of private investment in the US alone last year.

I’m not going to get as technical as my fellow bird, Michael, was in his earlier post, but I highly recommend his article for insight into why the widespread adoption of this technology isn’t without its downsides. I’m specifically focusing on the transformation of Google and the shift in its usage and prominence taking place in front of us at breakneck speeds. For an explanation of why I picked this topic, I think I’ll relay the short anecdote that prompted this article.

My Story of Searching

I took my first IT job in early 2021, but I had been tech-savvy and quite the gadget-head for much longer. After I joined that organization, I became responsible for training some front desk coworkers, and my job was to get them up-to-speed enough on technical matters during the slow season. One of the things we instilled in these trainees was the importance of Googling an issue before tagging a supervisor or trainer for additional help. I’m confident that step reads as a no-brainer, but I promise it’s easy to get lost in the weeds when you’re troubleshooting something. Recently, however, I found myself troubleshooting a relatively obscure issue with a client and had to resort to Google his specific error.

What followed next had me feeling insane; I was met with an AI summarization at the top of the Google search results that was completely wrong and wasn’t even about the issue I searched for. The sources linked in the summarization were Microsoft guides and forum posts about the correct program, but were otherwise unrelated. So, I resolved to pursue the search results and was immediately turned off to discover several similar-looking articles on websites I’d never heard of with generic headlines. They all, in my estimation, had hints of AI in their writing style—and they contained vague, non-specific troubleshooting steps (such as restarting your computer). Perhaps crazier, I may never know if those articles were written by an LLM (Large Language Model, i.e., ChatGPT) or real technicians or a combination of both. I didn’t have too much time to ruminate on this because at that moment, to my surprise, my client found a workaround… from ChatGPT.

The Search Problem

Quickly, I’m realizing that Google isn’t the tool it used to be. Even before the AI revolution, it was a robot attempting to sort search results, a system that was all too easily gamed by robots manipulating the tool to appear higher on the list. But now, Google is not even interested in being a helpful search engine; they’ve pivoted to being your answer machine—to telling you the truth rather than linking you to the possible answers. That risks a lot more than just being unable to find helpful answers to weird tech questions. To echo Dr. Hammond of CASMI from his article on the matter—“as we rely more on AI-driven answers, we risk losing our ability to engage with information critically”-Google might become the “curator of truth”.

This is becoming a problem for Google and its parent company very quickly. Just this month, digital news outlet The Verge reported on Penske Media suing Google for diverting traffic away from published works by promoting their in-house AI summaries. As author Terrence O’Brien noted, Penske is hardly the first company to take issue with Google’s use of their incredible influence. O’Brien cited lawsuits from Chegg and a coalition of publishers from Europe submitted earlier this year. For a quick reminder, Google was the most visited website in the world last month, per Similarweb, with a reported 83.9 billion visits across mobile and desktop. Second place is YouTube, as an aside, but an important note of the monolithic power Google (or, rather, Alphabet) has on the internet. Even the sources for this article I collected with Google’s help.

In Search of a Solution

I haven’t been shy in this rant of an article or in regular conversation to admit my concern with the LLM explosion we find ourselves living in, especially given how much I witness people personally and professionally take ChatGPT or Google AI Summaries as accepted truth. I don’t, however, intend to be all alarmist or doom and gloom. Quite the opposite. The truth is that Google and other major tech companies are always changing and innovating. Amazon used to sell books, and Facebook used to be a tool to help Harvard students connect. More than anything, I wish to communicate how crazy fast things are changing in the AI world. The landscape of the internet is changing rapidly, and the process of making sure technicians Googled something before asking for help is changing drastically, if those days aren’t ending altogether.

As technology continues to evolve in ways that are unpredictable to us, I would encourage you, readers, to be optimistic but mindful. The tools we use to interface with the world shape us, more so than ever, as our tools (like phones) become indispensable parts of us. It’s very feasible in my mind, as LLMs, so advanced they appear to be alive, find footholds in the home and workplace, that our tools might make us the kind of person we don’t intend on becoming if we aren’t aware. Just as improper social media adoption weakened our capacity for true connection, so might the uber-convenient and unregulated LLM usage weaken our capacity for being artistic and truthful people. It will take deliberate and measured steps to gain the benefit of these fantastic new tools while continuing to preserve those awesome qualities in ourselves, like art and truth.