The artificial intelligence war is on. The buzz surrounding ChatGPT-3 since its release to the general public has acted as an electroshock within the ecosystem of technology giants. The ability of the tool to formulate clear, synthetic, and structured answers with disconcerting ease, as well as its ability to generate textual content, have pushed OpenIA’s competitors to put forward their own projects.
It is now done for Google with the presentation of Bard, its AI-based conversational robot. Google Bard is both a response to ChatGPT, a preventive attack against Microsoft, which aims to integrate this technology into its tools, but also (and above all) a vision of what search engines could become in a few months or years. Focus.
What is Google Bard?
At the beginning of February, Google unveiled its own conversational AI, available exclusively in closed beta for the moment. Like ChatGPT, Google Bard is a chatbot that uses conversational learning models to provide answers to questions asked of it. Google has presented it as a new service in its own right, capable of interpreting a request and generating a complete and structured response in return.
Ever since Open AI made its innovative tool available to the general public, Google was expected to be on the ball. It was clear that the Mountain View company would not stand idly by in the face of the tsunami caused by ChatGPT – with its 100 million users in just two months, according to a recent study.
In the world of Big Tech, it has been rumored for a long time that Alphabet (Google’s parent company) was developing its own conversational agent powered by artificial intelligence.
But for Google, it was high time to come out of the woodwork, as Microsoft had already announced the integration of ChatGPT to its search engine, Bing, in the next few months.
The sudden popularity of ChatGPT has forced the American giant to show its teeth. Although Google Bard is still a promise, its announcement, through an official statement by CEO Sundar Pichai, as well as its presentation in Paris on February 8, were events. However, this is only a stop in a race that started several years ago.
Google and artificial intelligence
Google is not a neophyte in artificial intelligence. The firm has been developing projects around AI for six years, including (but not limited to) the R&D center opened in Paris in 2018, where the February 8 presentation was held. In 2017, the company already introduced Transformer, a deep-learning model used in the field of automatic language processing.
BERT and then MUM were built on this model: two algorithmic evolutions, integrated into the search engine, capable of better understanding natural language and search intentions, and of providing more sophisticated answers to complex requests, in a multitude of languages.
The irony is that the Transformer model is also used by… ChatGPT! (This is what makes Yann LeCun, the “Mr. AI” at Meta, say that the underlying technologies of this tool are not as innovative as they would have us believe).
In 2021, Google unveiled LaMDA (Language Model for Dialogue Applications), its revolutionary language model also built on Transformer. LaMDA has been trained to dialogue, but also to pay attention to the context in which its responses are formulated, paving the way for Google Bard.
How does Google Bard work?
Google Bard combines the breadth of global knowledge with the power, intelligence, and creativity of Large Language Models. Like ChatGPT, it relies on machine learning models trained on a huge corpus of data.
When a user asks a question or makes a request, the tool compiles a clear, concise, high-value-added answer, written by the user in his own words.
For Google, Bard is intended to simplify complex subjects and formulate answers that can be used by the greatest number of people. In his press release, Sundar Pichai gives this example: he asks the application to quote the discoveries made by the James Webb Space Telescope that he is likely to be able to explain to his 9-year-old child. The chatbot provides a three-point answer that is precise, useful, and well-adapted to the target.
(Source: Google)
(The more attentive among you will have noticed an error in Bard’s answer – but we’ll come back to that later).
What is interesting is that this example was not chosen at random. Open AI has made it clear that the database ChatGPT has trained on ends in the middle of 2021, so any information after that cannot be used by the tool in its answers (unless it has learned it by talking to people).
Since the James Webb Telescope was launched by NASA at the end of 2021, ChatGPT simply cannot answer a question about it. In doing so, Google points out a major difference between the two systems: unlike its counterpart, Bard is connected to the web, which allows it to tap into very recent data and bring users more relevant information.
The other unique feature of Google Bard is that it is powered by a “lighter” version of LaMDA, which is less resource intensive. This makes the technology easier to deploy on a large scale, as it can be used by a larger number of people simultaneously. For future integration with the search engine, this is an important detail.
Google Bard and the future of the search engine
This unveiling remains very elusive, especially since it is impossible, for the moment, to get our hands on Google Bard – only “trusted testers” will be asked by the firm to evaluate the tool. Moreover, Bard is already making mistakes and did so again, live, during its presentation in Paris (see here).
This factual error, which attributes to the JWST the first picture of a planet outside our solar system (while the real first image dates from 2004), shows that the tool is still in full development and that its integration to the search engine is not for soon.
However, this AI war gives a glimpse of what could be the future of search with the support of technologies like ChatGPT and Google Bard, especially by optimizing the quality of the enriched answers in the SERP.
It’s all about the tool’s ability to produce structured, synthetic content based on a wide range of information. While Google’s rich results are often limited to displaying excerpts from selected content on the Web or raw data, ChatGPT can reformulate this information to produce new content that can respond precisely to the user’s request.
This is the functionality that Microsoft is looking to implement in Bing, and this is what Google is also planning with Bard.
The image below gives an idea of what the SERP could become in the near future, boosted by artificial intelligence. Faced with a complex request that requires comparing different data, the engine would provide an answer synthesizing the information drawn from the web, written specifically to fit the need expressed.
(Source: Google)
Nevertheless, Google does not see the future of search through the prism of a single, absolute answer, as ChatGPT provides today with aplomb (and by committing huge blunders). Instead, the firm intends to make the two types of results coexist, on one side the text generated by the artificial intelligence, and on the other, the links referring to the most relevant content around the subject – as is the case today.
Some will say that this vision of the future is crucial for Google, which will do everything to maintain its business model based on advertising – and this is absolutely true – but it is also crucial for the free choice of Internet users. Which, in the context of a fake news outbreak, is not a trivial consideration.
After all, we all like to have a choice… and it would be desirable that the advent of AI in search engines does not have the perverse effect of directing (even more than now) the users’ thoughts toward the answers that the tool deems the most suitable!
Google Bard is still a promise. The company is still cautious, and this is understandable considering the many problems encountered by artificial intelligence applications over time (let’s remember that last year, Meta had to pull the plug on Galactica after three days of public access because the AI started producing inappropriate content).
The tool will have to be tested extensively and refined before being implemented anywhere. But what we can say is that the battle of the AI giants is just beginning.