Can ChatGPT overthrow Google?

All of the alarms went off last month at Google's headquarters.

All alarms went off at Google's headquarters last month.

The New York Times has devoted an entire article to the subject. According to them, 'Code Red' was activated in the company's highest levels.

Google is terrified, very terrified.

The reason?

Artificial intelligence has recently made a giant leap that could put Google's core business, search, in jeopardy.

The question becomes unavoidable.

Could we soon see the demise of one of the so-called "trillion-dollar" companies, and thus entire industries like SEO, SERPs, and, eventually, digital marketing?

Google, the original Internet monopoly, is extremely vulnerable.

Google is currently worth a whopping $1.13 trillion.

That's a one and a twelve. What's even more incredible is that Google will be worth nearly $2 trillion in November 2021.

This is a significant decrease, but it still places them as the world's fourth-largest company by market capitalization.

Of course, given its size, one would assume that its revenue is also massive, right?

In 2021, revenue is expected to be 256 billion dollars. Not too shabby.

But how much is it?

For comparison, this is more than Portugal's entire GDP for 2022.

In other words, Google's revenues in 2021 will be greater than what Portugal as a whole produced in a single year.

With those figures, it's impossible to be surprised by Google's exorbitant valuation.

Or should we?

To be honest, there is a catch to Google's business model.

Google's strength is not diversification.

It's impossible to deny that Google has some impressive numbers.

Consider the following monographic as a more visual representation:

If we look at Google's quarterly results for June 2022, we can see that the company earned 69.7 billion dollars.

Almost as impressive as their final profit of $16 billion, which represents a profit margin of 23%.

That's significant given that Amazon, one of the other big players, actually lost money that quarter.

Isn't this all amazing and wonderful?

However, if you look closely, you will notice something strange.

Or, at the very least, improvable.

Out of the 70 billion in revenue, 41 billion, or nearly 60%, comes from a single source: search advertising, an industry in which Google has a 92% market share.

And the problem is that this is precisely the market that AI has the potential to completely disrupt.

But how exactly?

AI models that can be transformed. But what exactly is that?

They aren't killer robots like in the movies, just a heads up.

ChatGPT and what's Next

You've probably heard or read a lot about ChatGPT recently.

A lot, painfully so.

But it's understandable; technology is incredible.

ChatGPT in 30 seconds

For those who are unfamiliar, OpenAI, a non-profit organisation founded by Elon Musk and Sam Altman, among others, has released the latest version of its chatbot, ChatGPT, which is powered by the largest transformer language model ever assembled, GPT-3.5, which has over 175 billion parameters.

In case you're wondering, chatbots are talking robots.

Actually, you've probably spoken to quite a few recently, especially when dealing with call centers, insurance companies, and pretty much any other type of customer service available today.

In all honesty, those chatbots are vexing and severely limited.

But not this one.

One could argue that it's the same animal but a completely different beast, but it's also a completely different animal to me.

ChatGPT can, among other things, respond to almost any question with extremely eloquent responses, code anything you want in a variety of programming languages, create entirely new bedtime stories, debug your own code, and so on.

It's so impressive that some believe it could be the world's first intelligent, sentient AI model.

Fuck no, is the short answer.

The issue with that statement is that it confuses high predictability with sentience.

A probabilistic monster

GPT, like any other neural network, is a probabilistic beast capable of predicting the next correct word in response to a sentence with astonishing accuracy, resulting in perfectly-elaborated sentences while sounding very human-like when interacting with it.

But being able to predict eloquent responses is one thing; being able to truly understand what they're responding to is quite another.

Don't be misled by sensationalist media; cutting-edge AI isn't sentient.

Although an armed uprising of human-killing robots appears unlikely today, there is something that ChatGPT is poised to disrupt from the start.

And that is Google's golden goose, Internet search.

Is Google Search and SEO doomed?

Unlike Google Search, ChatGPT saves you from having to scroll through endless pages of links while providing concise and direct responses.

This may lead to people querying these systems rather than searching on Google.

In that case, the appeal for businesses to promote themselves through Google could plummet, putting Google at risk.

However, Google is not the only one.

Thousands of marketing firms rely on their SEO expertise, or the ability to assist their clients in being 'found' on Google. This abrupt shift in customer journeys to use AI instead of Google could be disastrous for them.

Given how impressive ChatGPT is already, and the fact that a newer version of the underlying AI model, GPT-4, will be released early next year...

Are we willing to 'kill' Google?

Distinguishing between what is wrong and what is right

Developing a good search engine is not an easy task.

In order to provide the searcher with relevant but acceptable information, these algorithms must be constantly fine-tuned to avoid exposing fake news, racism, homophobia, or even plagiarism.

This effort appears to be an even greater challenge with AI.

Today, AI must determine three things:

Opacity

Data Bias

Avoiding making shit up

AI, the ultimate mystery

GPT-3.5, as previously described, has been fine-tuned for 175 billion parameters, requiring 800 GB of storage.

In layman's terms, this means that explaining how these models work and make decisions or predictions is nearly impossible.

As a result, if ChatGPT makes a racist remark, there is no way to explain why.

Data bias, model bias

Furthermore, because these models are not sentient but mathematical models that have learned to answer questions through brute force and massive amounts of data, they are extremely reliant on unbiased data sources and diverse teams of data engineers, which is a huge challenge in today's world.

Why is this a challenge? Here's an interesting fact for you.

According to Zippia data from September, 70% of computer programmers are men, and 69% of all programmers are white.

Even if the majority of them are not racist, this extremely skewed workforce is undoubtedly culturally biased, which is not ideal for AI models that aim to be universal and applicable to all of society.

Google Search is not going away... for the time being.

Despite this, I don't believe Google's reign as the supreme search engine is in jeopardy for the time being.

However, I believe that the future of search engines will be AI-led, so while today's AI models are limited and dangerous for widespread use, ChatGPT has shown us what the future undoubtedly looks like.

Fortunately for Google, it has its own large language model, LaMDa, and has undoubtedly taken note of what LLMs are capable of as a result of OpenAI.

Nonetheless, all of this demonstrates how disruptive AI will be. But not just for you and me, but also for the world's largest corporations.

A final word

The Internet contains far too much garbage.

In fact, quality content is difficult to come by, especially in Crypto or AI. It takes time and, more often than not, results in nothing.

But you know that if you want to be successful, you must stay ahead of the curve. Yes, the truth is painful.