Hello, join me as we look at the topic, Google Search is dead and the danger of LLMs. “Google Search is dead”, that’s a bold claim to make as Google search still works but there are things going on in the world of Internet Search that you should be aware of.

I’m sure that we all use the net so this is something that should in large parts concern us all. I’m hoping that I’m jumping the gun on this one but the handwriting is glaringly clear on the wall. It’s hard to escape it.

Google Search

The History of Search

What we now call the Internet started in 1993 and slowly but surely, websites began springing up all over the world.

However in an age of limited information, where most people still used Wikipedia as a physical book, finding information and the sites that held such information was incredibly difficult. We don’t even have to go that far back, as recently as 2006 after getting my first “GPRS-Enabled” phone as MTN called it, I had to copy my favourite sites on the back of my Commerce notebook. It was the most useless subject in SS1 anyways. Don’t judge me.

Anyway let’s go back to the 90s. With the early internet being so messy and disorganized, some very innovative people came up with the idea to curate and organize internet. The way one would do to a disorganized library, and then build a simple web program that would give the user access or an easy way to search for whatever they wanted.

This was how the idea of what we now know as Search Engines were born.

Google Search

Of course today, everyone and their dog knows that Google is the number one search engine. In fact, Google has become a euphemism for the term search. It would interest you to know that there were search engines around before Google. However Google somehow found a way to take over the market (over 90%) and have been the standard for how we search for stuff online for the last 15 years at least.

During this period millions of websites have been created and all of these websites have neatly organized and curated by Google.

Google Search Algorithm

It did not take long before several websites started competing for search terms (keywords) and the question of “which site should the user see first?” arose.

This is where the Google Search Algorithm comes in. This is a system by which not just Google, but other search engines and social media platforms can rank content and then choose the order in which to show them to the user.

With the rise of big data and deep learning AI models, algorithms can be configured to collect data on specific individuals and choose content to show them. This is why the Facebook feeds of two people can never be alike. I digress.

As our technology got better, websites multiplied. As the Internet grew, it became a legitimate virtual space for anything you wanted it to be. Do you want the Internet to be a community forum? Done! Do you want it be a gaming platform? Easy. A market place for buying and selling? Yes. A place for learning? Well e-learning is a thing now, isn’t it?

And through it all, Google grew and grew and grew. It also helped that they used Android as a Trojan horse to force their services down everyone’s throats.

Pull a random person off the street and ask them to name 3 search engines besides Google, I doubt if you can find anyone to be honest.

Large Language Models LLMs

Large Language Models (LLMs) and AI

So the story has been all nice and dandy so far, Jeff, how has search changed and how does it affect me? Get to that part.

Ah yes of course. Apart from being a search engine, Google is now many things. Top of the list is that Google is an Ad company. This means that they sell ad spaces on their search engines and on websites. They also dominate the ad market as well and made billions last year. Then a little something called LLMs showed up.

LLMs are AI language models that have been trained on large data sets and can repeat the information like a parrot if asked.

Google has had this technology (allegedly) but sat on it for years because their current search engine dominance and ad business was raking in billions yearly.

ChatGPT is born

When a company called Open AI figured out how to add Machine Learning to their LLM, Chat GPT was born. Chat GPT changed everything. No longer did you have to go searching through several pages and blogs/websites to find the information that you needed. Here was a bot that could provide all of the information that you needed with a few well worded instructions (prompts).

See: How to prompt chatbots (LLMs)

The game had changed overnight. If that wasn’t enough, Microsoft partnered with Open AI and Copilot was born. Microsoft tied Copilot with Bing and the flagging Search Engine suddenly had new life.

In the background, LLMs were springing up like mushrooms and luring people away from search engines. These LLMs were quasi search engines as they went straight to site that had the information and then reworked it for you in a way that fit your search intent.

Google went into panic mode and release Gemini. The LLM that they’ve been sitting on for years now. But it was too late. Chat GPT and Copilot had already captured the imagination of the general public.

With LLMs now popular, people would not have to visit Google (traffic). With less traffic, Google cannot show ads to readers and if they can’t, no one would pay Google to display ads.

Google’s March – May 2024 Update

So it happened that in March of 2024, Google unleashed a core update that decimated 80% of all websites, blogs etc. Inquisitive Universe was especially hit hard and has lost 80% of its traffic.

On May 5th, with the site reputation abuse update, Google also went after the big sites that survived the March to April cull. In the place of websites and blogs, Google has decided to rank user generated content (UGC) instead. This mostly includes content from forums like Reddit, Quora and Medium high on the SERPs whilst keeping a few big and trusted sites around.

If you search for something on Google, and then search for the same thing on Bing, DuckduckGo, Ecosia or Yahoo search, you’ll see far richer results on the latter than on Google. This is because Google has wiped out a lot of sites from appearing on their search engines.

Artificial Intelligence Generative AI

Search Generative Experience (SGE)

Google already has something big planned. It’s currently in Beta and it is called the Search Generative Experience. Google has already trained it’s LLM on all the data in the world and thus doesn’t need websites anymore. This is basically an AI search engine that just takes you straight to what you need to know, bypassing the blog or website that wrote it.

Google Search Generative Experience

At this point, it is easier to say that blogging or owning a website of any type is probably dead in the water.

As more and more people are choosing to engage with LLMs, search engines have quickly become obsolete. Relying on Google search is dead as LLMs seem to offer more benefits.

Benefits of LLMs

Before I get into all that, I wanna put this out. I don’t hate LLMs neither do I hold any bias against LLMs. On the contrary, they’ve made my life so much more easier.

  • They’ve made my research faster and have helped boost my writing output by drafting articles for me to rewrite to my taste.
  • They have also become virtual assistants to workers as well as companions to people who just need a friend.

There’s many benefits that LLMs can and do offer us, but it’s pitfalls are many and if they’re not mitigated, it could lead us into a worse place.

Pitfalls of LLMs

1. Poorly Written Content/Misinformation

One of the first ways it could do this is that it would engender laziness and chip IQ points off the general public. It has already began.

A lot of people are using Chat GPT to write term papers and projects without bothering to proofread them. A lot of bloggers even from grade A sites are writing with Chat GPT and do not even bother to edit or fact check them for accuracy before hitting publish.

In the last 12 months alone, the rate at which new sites are being created have tripled and many of these sites are filled with garbage AI posts. Imagine seeing a nice title and you click in only to read regurgitated and lifeless AI text that the author couldn’t even be bothered to reword.

This is one of Google’s justification for flattening the playing field claiming it was trying to eliminate AI spam sites.

2. Destruction of Blogs/Websites

The second one is destruction of websites and the ecosystems that have been built around them. A very good example of this is WordPress.

WordPress is the world’s most popular CMS and has a very active community of developers who optimize and make plugins for it. That’s not to mention to multitude of bloggers, website creators, online shop owners, forum creators etc. who rely on WordPress.

3. Loss of jobs

With LLMs stepping up into the limelight, a lot of these people, especially those who rely on the Internet for their livelihoods would have to look elsewhere or starve.

The job of SEOs, writers, web admins etc. are already under threat as a lot of them have been released from their jobs.

This would then take me into what I consider the most damning effect of the new order. The loss of millions of websites which would no doubt go under.

4. Lack of content to train LLMs

LLMs are like parrots, albeit highly advanced digital parrots. They do not have a mind of their own and they are certainly not capable of generating new information (at least not yet).

They rely heavily on preexisting data created by humans inorder to learn and regurgitate that data to someone else. With millions of websites gone, how would these LLMs do their jobs?

5. One sided information

Again, with many of these websites out of business, the ability to be able to read several articles from different people arguing for or against a point will be lost. In 2019/20, when marketers swamped Nairaland promoting their garbage, we were able to offer a differing and opposite message for the audience to read and choose.

But we could be heading into a future where differing and opposing voices as well as third party views on many topics could be gone and gone for good. I don’t think I need to tell you the inherent dangers of being fed information from only one perspective. It gets worse as LLMs can be programmed to have certain biases and could manipulate such information for the user who would accept it whole.

This is something I have tested myself. I took the text I typed here yesterday and fed it to Chat GPT, Copilot and Gemini. The first two analyzed it objectively but Gemini was very defensive of Google and urged me to take a more balanced approach instead of being critical of Google. I kid you not.

It is not beyond the owners of these LLMs to configure their bots to speak in favour of certain topics or bodies whilst being overly critical or dismissive of others. Tech giants are already configuring their algorithms to manipulate people on a massive scale. It is near certain that the same would be done to these LLMs.

6. Over dependence on LLMs

The last one on my list is that people could get so dependent on these LLMs that they couldn’t live without them. A day will come when LLMs will be monetized and walled off, people will become addicts, paying to be able access them.

The End

Microsoft did this whole thing best, having Copilot side by side with the Bing search engine. That way I can use Copilot to do a skeleton draft and then fact check it from an actual site. Yes LLMs and traditional Search can exist side by side.

That’s not to say that I’m an advocate for websites. Honestly, a lot of garbage sites did have to go. Especially those for Adsense, Ezoic, Adsterra, Mediavine sites that copied and scrapped data looking for money whilst offering no value. They needed to go and good riddance.

As it stands, Google Search is dead and LLMs seem to be the future. Yeah that’s what going on in the world of the Internet right now and how search is changing.

Please leave a comment if you have any difficulty and remember to: