Bard Google vs. Chat GPT Bing. A war of AI-powered search engines

Bard Google vs. Chat GPT Bing

The search engine war (fuelled by artificial intelligence) has begun!

We could even have added Baïdu - Ernie, which is also entering the intelligent conversational race, although there is a good chance that, as Baïdu is at present, its scope will remain solely Chinese, unlike the first 2 protagonists.

Google has been hegemonic for years, with a market share of almost 95% in Internet search worldwide, but it has been caught unawares by the arrival of Chat GPT, and above all by the agreement with Microsoft Bing, which links AI (Artificial Intelligence) to Internet search.

Google has also been working on the subject for years, but did not think that its 'weak' competitor would be so quick to strike back, and has been forced to retaliate as early as 2023.

What are the issues for search engines themselves? What is at stake for the referencing of websites and advertisers? 

We'll try to decipher them here.

Let's start by defining them 


What is CHAT GPT ( and BING + CHAT GPT ) ?

Chat GPT is a natural language model developed by OpenAI.

It is designed to understand and generate human language in a similar way to a human.

Based on a neural network architecture known as a "transformer", Chat GPT can process large quantities of text to learn to predict the next words in a text sequence.

Thanks to this, Chat GPT is capable of answering a variety of natural language questions and tasks, such as language translation, text understanding, text generation and much more.

In short, Chat GPT is a tool designed to understand and generate natural language, and it has been trained on large amounts of textual data to be able to understand and produce texts in several languages.

Bing, Microsoft's search engine and Google's challenger, has teamed up with Open AI to develop a more effective search, or at least one that complements search results from the web, in order to enrich them and try to provide more comprehensive results with a touch of artificial intelligence.


What is BARD from Google?

Bard is a natural language model developed by Google, which aims to improve the understanding and generation of natural language. Like Chat GPT, Bard is based on a neural network architecture called Transformer and uses massive training on text data to improve its performance.

However, there are a few differences between Bard and Chat GPT. Bard was designed with a more specific goal in mind, namely to improve the understanding of natural language search queries on Google. It is also optimised to run on mobile devices, which means it is lighter and faster than many other natural language models.

In short, Bard is a natural language model developed by Google that has been specifically designed to improve the understanding of natural language search queries on Google, particularly on mobile devices.


What are the challenges of AI-powered Internet research?

AI-powered Internet search presents a number of important issues that are worth examining. Here are some of these issues:

1 Quality and relevance of research results with AI :

The use of AI to improve Internet search has the potential to improve the quality and relevance of search results.

However, this can also lead to a monopolisation of information, where the large technology companies developing these technologies could influence search results according to their commercial interests.

Who will be able to control this bias for either Bing or Google? 

As long as we were dealing with site rankings, there was the potential for "commercial favouritism" ( see. SEA - Search Engine Adversised - ) but no opinion bias.

With the arrival of AI, who will guarantee that these biases, these allegations are objective and potentially contestable in some cases?

2. Data confidentiality :

AI-based research models require large amounts of data to operate effectively, which raises concerns about data confidentiality.

Users may not be aware of the amount of data they are sharing and how it is being used. Even less than they are today.

3. Algorithmic biases linked to artificial intelligence :

AI algorithms can be biased because of the data they have been trained to use.
If this data is biased, it can be reflected in search results, which can have negative consequences for users looking for accurate, objective information. 

Which ties in with point 1.

4. Impact on employment :

The use of AI to improve Internet searching could potentially have an impact on employment in research-related industries, including libraries, archives and museums.

In the same way, we can imagine that the production of human content will be increasingly robotised, and that the production of content for websites will themselves be generated more and more by machines, and less and less by humans.

5. Reliability of information sources :


While AI-powered web search can help improve the quality of search results, it can also make it harder to distinguish between reliable and unreliable sources of information.

AI algorithms can rank results based on popularity or engagement, rather than the reliability or accuracy of the information.

6. Excessive personalisation of search results in SERPs :

AI algorithms can use personal data to personalise search results based on user preferences and behaviour.

While this can be useful for some searches, it can also lead to over-personalisation and a filter bubble, where users are only exposed to information that matches their opinions and interests.

7. Cost and access to resources :

AI-based research models require large amounts of data, computing and IT infrastructure to operate effectively.

This can make the use of these technologies expensive for small businesses or not-for-profit organisations that do not have the resources to afford these costs.

8. Regulations and liability :

As with many other AI applications, AI-powered web search raises questions of regulation and liability.

The companies developing these technologies must be responsible for the way they use the data, guarantee the transparency of the algorithms and search results, and comply with data protection and privacy laws and standards.



In short, AI-powered web search can offer significant benefits in terms of the quality, relevance and personalisation of search results, but it also raises a number of significant concerns in terms of data protection, reliability of information sources, regulation and costs.

It is therefore important that these issues are addressed proactively to maximise the potential benefits of these technologies while minimising the risks.


Google Bard vs CHAT GPT. Who will win?

It's hard to say at this stage, but one thing is certain: the cards could be reshuffled depending on the relevance that Internet users find in one or both search engines.

We can even assume quite easily that, initially, searches are carried out AND on one side AND on the other to observe which of the two provides the best information, and either eventually keep both in use, or keep only the one of the two whose answers seem to be the most relevant. 

In any event, there could be big losers indirectly, such as Wikipedia, and the providers Google and Microsoft themselves, if they are unable to retain sponsored links (and therefore the advertising revenue linked to these sponsored links).

We can therefore imagine that, to avoid killing the goose that lays the golden eggs, the results of intelligent research will be a clever addition to the SERP without taking up all the space, which is itself limited.


The possible consequences of AI-powered search engines

Did Prabhakar Raghavan, vice-president of US giant Google, announce the arrival of Bard too soon?

A simple test error by Bard caused Google to lose $100 billion in market capitalisation in one day...

We can therefore already foresee some rather brutal and violent movements over the coming months.

Internet users' uses are changing thanks to / because of AI

Until then, Internet users were actively searching for sites on Google, Bing and other sites. 
From now on, they'll come to us for substantiated answers, just as if they were consulting a live Wikipedia.

Search engines will therefore need to be able to distinguish between searches for information purposes and those for transactional purposes.

Search intent is therefore going to determine the face of the SERP, and we're going to have to manage the space taken up by "discussion", which will have to continue to coexist with clickable links, especially on smartphones.

It would appear that Bing - Chat GPT already includes micro-links within the enriched conversational content.

Website rankings evolving thanks to / because of AI

As a result, with space becoming more expensive (because it's rarer), what will happen to sponsored links and website SEO rankings?

There's a good chance that this will complicate the work of SEO agencies, and/or that it will offer them other possibilities for positioning.

It is also very likely that Bing will take back market share from Google, and that the work hitherto concentrated mainly on Google will be shared out a little more between the two search engines. 

Loss of market share for Google (and gains for Bing)

With the emphasis on speed, Chat GPT and Bing are in the process of introducing new features, notably earlier than Bard Google, which was taken by surprise.

Now, in what proportions and up to what point?

Everything will depend on the perceived quality of the two systems by Internet users, and the perceived relevance of the search results, which will be a mix of transactional and informative.

But it's a safe bet that Bing will be back in the game after years of drought, and so a new generation of engine.

L'user experience and the quality of the responses will be at the heart of the debates on these market share gains or losses (and/or losses for Google).

Consequences for transactional searches and e-commerce sites

What will become of Google Shopping in a world of restricted SERPs?

Either, as we are already seeing with Bing - Chat GPT, the conversational part will be relegated to a dialogue box to the right of the classic results, or it will become intertwined with the URL ranking... and then we could fear the worst when it comes to the place reserved for sponsored ad thumbnails...

Some sites could potentially lose millions of euros as a result E-Commerceor, failing that, to allow only the biggest players to grab the best remaining places at the expense of the "smallest" players.  

Consequences for the SEO profession and SEO agencies 

Both a risk and an opportunity, the content of websites could be replaced by the beginnings of fashionable answers. Featured Snippets "further expanded and improved 

In some cases, you no longer need to click on the link to the website concerned, as the response will already have been produced and obtained on the Google or Bing SERP.

However, there are two assumptions that can be made:

The first is that search engines will still need to be constantly fed with fresh, innovative content (produced by the news sites themselves - media, blogs, etc.).

The second is that the work of SEO agencies, which was essentially concentrated on Google, should now be shared out a little more between Bing and Google.

One thing is certain, end customers will need advice to ensure that they continue to rank well, with the added touch of AI. 

Optimize 360 is proud to promote its customers within the two enhanced systems 

Consequences for advertisers 

As mentioned above with Google Shopping, a new version of the SERP could result in fewer places available for sponsored links. Google AdsBing Ads etc.

We can therefore expect an increase in advertising bids, in order to continue to commercially promote websites in a restricted space of expression. 


In conclusion, the future is exciting with the arrival of conversational robots, and this new experience linked to the OpenAI language model, itself soon to face competition from LaMDA (a family of conversational neural language models developed by Google). 

Keep up to date with our latest developments.

This article will be updated as developments occur in 2023 and beyond.

Our other articles on EMS and AI

blank
About the author

Frédéric POULET is the founder of the SEO Agency Optimize 360 After 20 years in large groups, he founded Optimize 360 to put the know-how of large groups to work for SMEs.
blank Digital Performance Accelerator for SMEs

ContactAppointments

en_GBEN