What is SEO? Google AI Reverse Engineering (2022)

Search engine optimization (SEO) is the process of optimizing on-page and off-page factors that affect how high a website ranks for a particular keyword. This is a multi-faceted process that includes optimizing page load speed, generating a link building strategy, as well as learning how to reverse Google’s AI using computational thinking.

Computational thinking is an advanced type of analysis and problem-solving technique that computer programmers use when writing code and algorithms. Computational thinkers will seek ground truth by breaking down a problem and analyzing it using first-principles thinking.

Since Google isn’t giving out their secret sauce to anyone, we’ll rely on computational thinking. We’ll go through some key moments in Google’s history that shaped the algorithms used, and we’ll learn why this matters.

How to Create a Mind

We will begin with a book published in 2012 called “How to Create a Mind: The Secret of Human Thought Revealed” by renowned futurist and inventor Ray Kurzweil. This book dissected the human brain, breaking down the ways it works. We learn from the ground up how the brain trains itself using pattern recognition to become a prediction machine, always working to predict the future, even predicting the next word.

How do people recognize patterns in everyday life? How are these connections formed in the brain? The book begins with understanding hierarchical thinking, this is understanding a structure that is composed of different elements that are arranged in a pattern, this arrangement then represents a symbol such as a letter or a sign, and then this is further arranged into a more advanced pattern for example a word, and finally a sentence. Ultimately, these patterns form ideas, and these ideas are transformed into the products that humans are responsible for building.

By emulating the human brain, revealed is a path to create an advanced AI beyond the current capabilities of the neural networks that existed at the time of publication.

The book was a blueprint for creating a scalable AI by vacuuming the world’s data, and using its multi-layered pattern recognition processing to analyze text, images, audio and video. A system optimized for scale-up due to the advantages of the cloud and its parallel processing capabilities. In other words, there would be no maximum on data input or output.

This book was so pivotal that soon after its publication, author Ray Kurzweil was hired by Google to become director of engineering focusing on machine learning and language processing. A role that fit perfectly with the book he had written.

It would be impossible to deny how influential this book was to the future of Google and how they rank websites. This AI book should be required reading for anyone who wants to become an SEO expert.

DeepMind

Launched in 2010, DeepMind was a hot new startup that used a revolutionary new type of AI algorithm that took the world by storm, it was called reinforcement learning. DeepMind described it best as:

“We present the first deep learning model to successfully learn control policies directly from high-dimensional sensory inputs using reinforcement learning. The model is a convolutional neural network, trained with a variant of Q-learning, whose input is raw pixels and whose output is a value function which estimates future rewards.”

By combining deep learning with reinforcement learning, it became a deep reinforcement learning system. In 2013, DeepMind used these algorithms to rack up wins against human players on Atari 2600 games – and this was achieved by mimicking the human brain and how it learns from training and repetition.

Similar to how a human learns through repetition, whether it’s kicking a ball or playing Tetris, AI will also learn. The AI’s neural network tracked performance and would gradually improve itself, resulting in stronger feature selection in the next iteration.

DeepMind was so dominant in its technological edge that Google had to buy access to the technology. DeepMind was acquired for more than $500 million in 2014.

After the acquisition, the AI ​​industry witnessed successive breakthroughs, the kind not seen since May 11, 1997, when chess grandmaster Garry Kasparov lost the first game of a six-game match against Deep Blue, a chess-playing computer developed by researchers at IBM.

In 2015, DeepMind refined the algorithm to test it on Atari’s suite of 49 games, and the machine beat human performance on 23 of them.

That was just the beginning, later in 2015 DeepMind began focusing on AlphaGo, a program with the stated goal of defeating a professional Go world champion. The ancient game of Go, first seen in China around 4,000 years ago, is considered to be the most challenging game in human history, with its potential 10,360 possible moves.

DeepMind used supervised learning to train the AlphaGo system by learning from human players. Soon after, DeepMind made headlines after AlphaGo beat Lee Sedol, the world champion, in a five-game match in March 2016.

Not to be outdone, in October 2017 DeepMind released AlphaGo Zero, a new model with the key differentiator being that it required zero human training. Since it required no human training, nor did it require any labeling of data, the system essentially used unsupervised learning. AlphaGo Zero quickly outperformed its predecessor, as described by DeepMind.

“Previous versions of AlphaGo were originally trained on thousands of human amateur and professional games to learn how to play Go. AlphaGo Zero skips this step and learns to play simply by playing against itself, starting with completely random play. In doing so, it quickly surpassed human-level play and defeated the previously published champion-defeating version of AlphaGo by 100 games to 0.”

Meanwhile, the SEO world was hyper-focused on PageRank, the backbone of Google. It begins in 1995, when Larry Page and Sergey Brin were Ph.D. students at Stanford University. The duo began collaborating on a new research project nicknamed “BackRub”. The goal was to rank web pages to a measure of importance by converting their backlink data. A backlink is simply any link from one page to another, like this link.

The algorithm was later renamed PageRank, named after both the term “web page” and co-founder Larry Page. Larry Page and Sergey Brin had the ambitious goal of building a search engine that could power the entire web using backlinks.

PageRank Dominates Headlines

SEO professionals immediately understood the basics of how google calculates a quality ranking for a website using PageRank. Some savvy black hat SEO entrepreneurs took it a step further and understood that to scale content, it might make sense to buy links instead of waiting to acquire them organically.

A new economy emerged around feedback loops. Eager website owners who needed to influence search engine rankings would buy links, and in return, those desperate to make money from websites would sell them links.

The sites that bought links often invaded Google over the established brands.

Ranking using this method worked very well for a long time – until it stopped working, probably around the same time machine learning took off and solved the underlying problem. With the introduction of deep reinforcement learning, PageRank would become a ranking variable, not the dominant factor.

Now the SEO community is divided on link buying as a strategy. I personally believe that link buying produces suboptimal results and that the best methods of acquiring backlinks are based on variables that are industry specific. A legitimate service that I can recommend is called HARO (Help a Reporter). The opportunity at HARO is to obtain backlinks by fulfilling media requests.

Established brands never had to worry about acquiring links as they had the benefit of time working in their favor. The older a site is, the more time it has had to accumulate quality backlinks. In other words, a search engine ranking was highly dependent on the age of a website, if you calculate using the metric time = backlinks.

For example, CNN will naturally receive backlinks to a news article because of its brand, trust, and because it was listed high to begin with – so naturally it got more backlinks from people researching an article and linking to the first search result they found.

This means that higher ranked websites organically received more backlinks. Unfortunately, this meant that new sites were often forced to abuse the backlink algorithm by turning to a backlink marketplace.

In the early 2000s, buying backlinks worked remarkably well and was a simple process. Link buyers bought links from high-authority sites, often site-wide footer links, or perhaps on a per-article basis (often disguised as a guest post), and sellers desperate to monetize their sites were happy to oblige – unfortunately, often at the sacrifice of quality.

Eventually, Google’s talent pool of machine learning engineers realized that coding search engine results by hand was futile, and much of PageRank was handwritten coding. Instead, they understood that AI would eventually be responsible for calculating the rankings entirely with little to no human intervention.

To stay competitive, Google uses all the tools in its arsenal, and this includes deep reinforcement learning – the most advanced type of machine learning algorithm in the world.

This system layered on top of Google’s acquisition of MetaWeb was a game changer. The reason the MetaWeb acquisition in 2010 was so important is that it reduced the weight Google placed on keywords. Context was suddenly important, this was achieved by using a categorization methodology called “entities”. As Fast Company described:

When Metaweb figures out which device you’re referring to, it can return a set of results. It can even combine entities for more complex searches—“actresses over 40” might be one entity, “actresses living in New York City” might be another, and “actresses with a movie currently playing” might be another . “.

This technology was rolled into a major algorithm update called RankBrain that launched in spring 2015. Focusing on understanding context versus being purely keyword-based, RankBrain would also consider environmental contexts (eg the searcher’s location) and extrapolate meaning where there had been none before. This was an important update especially for mobile users.

Now that we understand how Google uses these technologies, let’s use computational theory to speculate how it’s done.

What is Deep Learning?

Deep learning is the most widely used type of machine learning – It would be impossible for Google not to use this algorithm.

Deep learning is significantly influenced by how the human brain works, and it attempts to mirror the brain’s behavior in how it uses pattern recognition to identify and categorize objects.

For example, if you see the letter a, your brain automatically recognizes the lines and shapes and then identifies it as the letter a. The same applies to the letters ap, your brain automatically tries to predict the future by coming up with potential words like app or apple. Other patterns may include numbers, road signs, or identifying a loved one in a crowded airport.

You can think of the connections in a deep learning system as similar to how the human brain operates with the connection of neurons and synapses.

Deep learning is ultimately the term for machine learning architectures that join together many multilayer perceptrons so that there is not just one hidden layer, but many hidden layers. The “deeper” the deep neural network is, the more sophisticated patterns the network can learn.

Fully connected networks can be combined with other machine learning capabilities to create different deep learning architectures.

How Google Uses Deep Learning

Google spiders the world’s websites by following hyperlinks (think neurons) that connect websites to each other. This was the original methodology that Google used from day one and is still in use today. Once websites are indexed, various types of AI are used to analyze this treasure trove of data.

Google’s system marks the web pages according to various internal calculations, with only minor human input or intervention. An example of an intervention could be the manual removal of a specific URL due to a DMCA takedown request.

Google engineers are known to frustrate attendees at SEO conferences, and this is because Google executives can never properly articulate how Google operates. When asked why certain sites don’t rank, it’s almost always the same poorly articulated answer. The response is so frequent that participants often state in advance that they have committed to creating good content for months or even years on end without positive results.

Predictably, website owners are instructed to focus on building valuable content – ​​an important component, but far from comprehensive.

This lack of response is because the managers are unable to properly answer the question. Google’s algorithm operates in a black box. There’s input, and then output – and that’s how deep learning works.

Now let’s go back to a ranking penalty that negatively affects millions of websites, often without the website owner’s knowledge.

PageSpeed Insights

Google is not often transparent, PageSpeed ​​​​Insights is the exception. Sites that fail this speed test will be sent into a penalty box for slow loading – especially if mobile users are affected.

What is suspected is that at some point in the process there is a decision tree that analyzes fast sites, versus slow loading (PageSpeed ​​Insights failed) sites. A decision tree is essentially an algorithmic approach that divides the data set into individual data points based on different criteria. The criteria can be to negatively affect how high a page ranks for mobile versus desktop users.

Hypothetically, a penalty could be applied to the natural ranking score. For example, a site that ranks #5 without penalty could have a -20, -50, or some other unknown variable that would drop the ranking to #25, #55, or some other number chosen by the AI.

In the future, we may see the end of PageSpeed ​​Insights, as Google becomes more confident in AI. This current intervention on speed by Google is dangerous as it potentially eliminates results that would have been optimal and it discriminates against the less tech savvy.

It’s a big ask to require anyone running a small business to have the expertise to successfully diagnose and correct speed test problems. A simple solution would be for Google to simply release a speed optimization plugin for wordpress users, as wordpress powers 43% of the internet.

Unfortunately, all SEO efforts are in vain if a website fails to pass Google’s PageSpeed ​​Insights. The effort is nothing less than a website disappearing from Google.

How you pass this test is an article for another time, but you should at least confirm if your site passes.

Another important technical metric to worry about is a security protocol called SSL (Secure Sockets Layer). This changes the URL of a domain from http to https, ensuring the secure transfer of data. Any site that does not have SSL enabled will be penalized. While there are some exceptions to this rule, online shopping and financial sites will be the hardest hit.

Low-cost web hosts charge an annual fee for SSL implementation, while good web hosts like Siteground issue SSL certificates for free and integrate them automatically.

Meta Data

Another important element of the website is the Meta Title and Meta description. These content fields have an outsized order of meaning that can contribute as much to the success or failure of a page as the entire content of the page.

This is because Google has a high probability of choosing the meta title and meta description to appear in the search results. And that’s why it’s important to fill in the meta title and meta description fields as carefully as possible.

Alternatively, Google may choose to ignore the meta title and meta description and instead automatically generate data that it predicts will result in more clicks. If Google poorly predicts which title to automatically generate, this will contribute to fewer clicks for searchers and consequently contribute to lost search engine rankings.

If Google believes that the included meta description is optimized to receive clicks, it will display it in the search results. If this fails, Google pulls a random piece of text from the website. Often Google chooses the best text on the page, the problem is that this is the lottery system and Google is consistently bad at choosing which description to choose.

Of course, if you think the content on your page is really good, it sometimes makes sense to let Google choose the optimized meta description that best matches the user search. We will not choose any meta description for this article as it is rich in content and Google will likely choose a good description.

Meanwhile, billions of people are clicking on the top search results – This is the human-in-the-loop, Google’s latest feedback mechanism – And this is where reinforcement learning starts.

What is Reinforcement Learning?

Reinforcement learning is a machine learning technique that involves training an AI agent through repetition of actions and associated rewards. A reinforcement learning agent experiments in an environment, takes actions, and is rewarded when the correct actions are taken. Over time, the agent learns to take the actions that will maximize the reward.

The reward can be based on a simple calculation that calculates the amount of time spent on a recommended page.

If you combine this methodology with a Human-in-the-loop subroutine, this will sound a lot like existing recommendation engines that control every aspect of our digital lives like YouTube, Netflix, Amazon Prime – And if it sounds like how a search engine should work you are right.

How Google Uses Reinforcement Learning

Google’s flywheel improves with each search, humans train the AI ​​by choosing the best result that best answers their search, and the similar search from millions of other users.

The reinforcement learning agent continuously works to improve itself by reinforcing only the most positive interactions between search and delivered search results.

Google measures how long it takes for a user to scan the results page, the URL they click on, and they measure the time spent on the visited site, and they record the return click. This data is then aggregated and compared for each site that offers a similar data match or user experience.

A site with a low retention rate (time spent on site) is then fed the reinforcement learning system with a negative value, and other competing sites are tested to improve the rankings offered. Google is objective, provided there is no manual intervention, Google eventually provides the desired search results page.

The users are the people who provide Google with free data and become the final component of the deep reinforcement learning system. In exchange for this service, Google offers the end user an opportunity to click on an advertisement.

The non-monetized ads act as a secondary ranking factor, providing more data about what makes a user want to click.

Google essentially learns what a user wants. This can be loosely compared to a recommendation engine of a video streaming service. In that case, a recommendation engine will feed users content that is targeted to their interests. For example, a user who usually enjoys a stream of romantic comedies might enjoy some parodies if they share the same comedians.

How Does this Help SEO?

If we continue with computational thinking, we can assume that Google has trained itself to deliver the best results, and this is often achieved by generalizing and catering to human biases. Indeed, it would be impossible for Google’s AI not to optimize results that accommodate these biases, if it did, the results would be suboptimal.

In other words, there is no magic formula, but there are some best practices.

It is the SEO practitioner’s responsibility to recognize the biases that Google is looking for that are specific to their industry – and to incorporate those biases. For example, someone searching for election results without specifying a date is most likely searching for the most recent results – this is a recency bias. Someone searching for a recipe most likely doesn’t need the latest site, and may actually prefer a recipe that has stood the test of time.

It is the SEO practitioner’s responsibility to offer visitors the results they are looking for. This is the most sustainable way to rank in Google.

Website owners need to stop targeting a specific keyword with the expectation that they can deliver whatever they want to the end user. The search result must match the user’s needs exactly.

What is a bias? It could be having a domain name that looks high authority, in other words does the domain name match the market you serve? Having a domain name with the word India in it can discourage US users from clicking on the URL, due to a nationalistic bias of trusting results coming from the user’s country of residence. Having a one-word domain can also give the illusion of authority.

The most important bias is what does a user want to match their search? Is it an FAQ, a top 10 list, a blog post? This must be answered, and the answer is easy to find. You just need to analyze your competitors by performing a Google search in your target market.

Black Hat SEO is Dead

Contrast this with Black Hat SEO, an aggressive method of website ranking that utilizes devious SPAM techniques, including backlink buying, backlink spoofing, website hacking, auto-generation of social bookmarking on a large scale, and other dark methods used via a network of black hat tools.

Tools that are often reused and resold on various search engine marketing forums, products with almost no value and little chance of success. Currently, these tools enable merchants to become wealthy while offering minimal value to the end user.

This is why I recommend leaving Black Hat. Focus SEO on viewing it through the lens of machine learning. It’s important to understand that every time someone skips over a search result to click on a result buried underneath, it’s the human-in-the-loop working with the deep reinforcement learning system. The human helps the AI ​​with self-improvement, and gets infinitely better as time progresses.

This is a machine learning algorithm that has been trained by more users than any other system in human history.

Google handles an average of 3.8 million searches per minute worldwide. That comes to 228 million searches per hour, 5.6 billion searches per day. That’s a lot of data, and this is why it’s stupid to try black hat SEO. It is foolish to assume that Google’s AI is going to remain stagnant and the system uses the law of accelerating returns to exponentially improve itself.

Google’s AI is becoming so powerful that it is conceivable that it could eventually become the first AI to reach Artificial General Intelligence (AGI). An AGI is an intelligence capable of using transfer learning to master one field and then applying that learned intelligence across multiple domains. While it may be interesting to explore Google’s future AGI efforts, it should be understood that once the process starts it is difficult to stop. This is of course speculating towards the future as Google is currently a type of narrow AI, but that is a topic for another article.

Knowing that this spending one more second on black hat is a fool’s errand.

White Hat SEO

If we accept that Google’s AI will continuously improve itself, we have no choice but to give up trying to outsmart Google. Instead, focus on optimizing a website to optimally give Google what it’s specifically looking for.

As described, this involves enabling SSL, optimizing page load speed and optimizing the meta title and meta description. To optimize these fields, the Meta Title and Meta Description must be compared to competing sites – Identify the winning elements that result in a high click-through rate.

If you optimized to get clicked, the next milestone is to create the best landing page. The goal is a landing page that optimizes user value so much that the average time spent on the page outperforms similar competitors vying for top search engine results.

Only by offering the best user experience can a website increase in ranking.

So far, we’ve identified these metrics as the most important:

The landing page is the most difficult element when competing against the world. The landing page must load quickly, and must serve everything expected, then surprise the user with more.

Final Thoughts

It would be easy to fill another 2,000 words describing other AI technologies that Google uses, as well as digging deeper into the rabbit hole of SEO. The purpose here is to direct attention to the most important calculations.

SEO partitioners are so focused on gaming the system that they forget that the most important element of SEO is ultimately providing users with as much value as possible.

One way to achieve this is to never allow important content to become obsolete. If in a month I think of an important contribution, it will be added to this article. Google can then identify how fresh the content is, matched with the history of the page that provides value.

If you are still worried about getting backlinks, the solution is simple. Respect the visitors’ time and give them value. The backlinks will come naturally, as users will find value in sharing your content.

The question then turns to the website owner as to how to provide the best user value and user experience.

IBM has been a leader in artificial intelligence since the 1950s. The company’s core offerings are IBM Watson, an AI-based cognitive service, AI software as a service, and scale-out systems designed to deliver cloud-based analytics and AI services.

Is SEO a boring job?

SEO: It’s Boring! The bottom line is that SEO takes time. It is ACCUMULATIVE. There is no quick fix to reach number 1 and stay there, no matter what the ‘gurus’ tell you. If you try to take shortcuts by “stuffing keywords”, creating funky backlinks, or spamming other people’s sites, Google will PANALIZE you anyway.

Is SEO fun for jobs? It is the broadest field within marketing. As a result, there is so much to learn and discover that a career in SEO is both challenging and rewarding, exciting and interesting.

Is SEO a stressful job?

Any job can be stressful based on the type of performance required by your company. It is the same with SEO related jobs. However, it is not as stressful as IT or other programming related jobs.

Is SEO worth it as a career?

Search Engine Optimization Career Opportunities ‘Yes. SEO is a good career option as it is among the most sought after careers in digital marketing. There are several organizations around the world that hire SEO professionals to generate better content and thus produce more leads.

Is SEO a hard job?

SEO isn’t that hard to learn, but it can be confusing and overwhelming to get started. Learning SEO means learning about a long list of individual digital marketing strategies, which can feel a bit like adding new weapons to your arsenal as you learn how to use them.

How many months it will take to learn SEO?

If you can learn SEO for a couple of hours every day, you can master the basics of SEO within 4-8 weeks and land your first SEO job in 3-6 months. If you can learn SEO full time, you can master the basics yourself within 1-2 weeks.

.

Will SEO be replaced by AI?

AI will completely change SEO as we know it. From optimization to link building, it will significantly affect all aspects of SEO. As artificial intelligence (AI) becomes more sophisticated, search engine optimization will need to adapt.

Will AI replace SEO writers? Yes, it may be able to reproduce our tone of voice and conversational style, but it simply cannot think the way we do and form its own opinions. It has no emotional intelligence. Our experience, research, training and social interactions are what help us shape our opinions and inspire others.

Will SEO exist in 10 years?

Speculation abounds, but they do not always provide a reliable forecast of what is to come. However, one thing is certain: SEO will be around for a long time. As long as search engines exist and internet users continue to use search terms and phrases to find what they are looking for, the search engine business will continue.

How long will SEO last?

Most professionals expect to see results in as little as 2 months, but SEO can take as much as 12 months to work. While every company’s SEO strategy is different, most businesses can expect to see significant results within 6 to 12 months.

Does SEO have a future?

SEO is only going to grow rapidly in the future. Just make sure to follow SEO best practices like creating useful content, having a faster website, and following Google’s guidelines to get better rankings.

Is SEO still relevant in 2022?

The fact that SEO works just fine even in 2022 as a method of delivering improved commercial results for business websites and their owners, combined with the need for continuous research into what works, makes it still relevant as a digital marketing method, but even more so as a valuable service and a rewarding…

Will SEO be replaced?

Will SEO still be relevant in 2022? Yes of course. Although some SEO tactics that were effective in the past have stopped working, SEO continued to evolve. Constantly reinventing itself to try to better match user intent, reducing spammy, ineffective tactics to improve.

Will SEO exist in 10 years?

Speculation abounds, but they do not always provide a reliable forecast of what is to come. However, one thing is certain: SEO will be around for a long time. As long as search engines exist and internet users continue to use search terms and phrases to find what they are looking for, the search engine business will continue.

How long will SEO last? Most professionals expect to see results in as little as 2 months, but SEO can take as much as 12 months to work. While every company’s SEO strategy is different, most businesses can expect to see significant results within 6 to 12 months.

Is SEO a long term career?

And also, the domain you work in changes your task and most of the time you learn new ways of SEO. There are different search engines and all the most popular search engines are constantly changing their algorithms to display results. So SEO is definitely not a bad idea in the long run.

Is SEO stressful job?

Any job can be stressful based on the type of performance required by your company. It is the same with SEO related jobs. However, it is not as stressful as IT or other programming related jobs.

Is SEO long term job?

Your SEO strategy for growth is a long-term investment, but you should also think about a long-term SEO strategy. If you’ve actually implemented a growth SEO strategy of some kind, hopefully it’s been a good move for your marketing and branding success.

Does SEO have a future?

SEO is only going to grow rapidly in the future. Just make sure to follow SEO best practices like creating useful content, having a faster website, and following Google’s guidelines to get better rankings.

What is the biggest threat of AI?

Existential risk from artificial general intelligence is the hypothesis that significant advances in artificial general intelligence (AGI) could result in human extinction or some other irreversible global catastrophe.

What is the biggest fear of AI? “Many fear that AI will make bad decisions. This fear is often very broad from a technical perspective, but it always boils down to people thinking that the decision “just isn’t right,” says Jeff McGehee, director of engineering at Very.

What is the biggest problem in AI?

One of the biggest problems with artificial intelligence is that the sophisticated and expensive processing resources needed are out of reach for most businesses. In addition, they lack access to the expensive and scarce AI expertise required to effectively utilize these resources.

What are the major threats to artificial intelligence?

Is artificial intelligence a threat? The technology community has long discussed the threats posed by artificial intelligence. Automation of jobs, the spread of fake news, and a dangerous arms race with AI-powered weapons have been suggested as some of the biggest dangers AI poses.

What are 4 risks of artificial intelligence?

What are the risks of artificial intelligence?

• Lack of traceability for AI implementation. …
• Introduces program bias into decision making. …
• Data source and breach of privacy. …
• Black Box Algorithms and Lack of Transparency. …
• Unclear legal responsibility.

What are 4 risk of AI?

Automation of jobs, the spread of fake news, and a dangerous arms race with AI-powered weapons have been suggested as some of the biggest dangers AI poses. Destructive superintelligence – aka artificial general intelligence that is created by humans and escapes our control to wreak havoc – is in a category of its own.

What are the 3 major AI issues?

AI presents three main areas of ethical concern to society: privacy and surveillance, bias and discrimination, and perhaps the deepest, most difficult philosophical question of our time, the role of human judgment, said Sandel, who teaches a course in moral, social. , and political implications of new technology.