Open AI’s AI-powered dialog tool ChatGPT has been in the news since its release, to the point that we’re talking about directly competing with Google and predicting the next death of the Mountain View company. However, all of this falls within the normal order of search tool development. And ChatGPT is still far from competing with the leading engine…
In the past few weeks, Open AI’s ChatGPT tool has been making headlines with its rather impressive question-and-answer system (even in French) powered by artificial intelligence. The system certainly has its limitations, but it remains efficient, especially informative, for a certain number of requests.
Here are two examples of questions for which the ChatGPT answer is of fairly good quality:
questions about SEO and Google, Replies from ChatGPT. Source: Abundance
But it must be noted that the answers are often riddled with inaccuracies, even errors. Here are a few examples among many others, with explanations in the legend:
Question about Spirou, created by Rob-vel and not by Franquin. A bellhop is not really a valet. Spip is a squirrel and not a hamster. The other designers didn’t wait for Franquin’s death to pick up the character. Etc. Source: abundance
Question about Iznogut : Here Goscinny is the screenwriter and not the designer (although he was before he was a screenwriter). But for Iznogoud, he is clearly the screenwriter and not the designer. Jean Tabary is French (although born in Stockholm) and not Franco-Belgian. Source: Abundance
Question about Abraracourcix : The wife of the chief of the Gallic village is called Bonemine and not Bonnaire. Source: Abundance
question about wine Kleven : No mention at the beginning of Heiligenstein, cradle of the Klevener since 1742 u unique City that produces this wine. Heiligenstein is also in the Bas-Rhin and not in the Haut-Rhin. Etc. Source: abundance
We see from these few examples (which are not intended to be a scientific study of the reliability of ChatGPT, but which could be repeated almost endlessly) that this tool can hardly be used without rejecting the result on other sources. Reliability is clearly non-existent (yet)… Ask ChatGPT a question on a topic you know well and you’ll recognize it instantly. It is unthinkable today to use ChatGPT as a Lambda search engine for daily information research.
Still, the tool is notable for the form of answers it returns (structured and long, descriptive sentences, with no spelling mistakes) and in that it shows promise for what will come when the many current errors are corrected. Today it can be assumed that 90% of the content of the answers given is quite reliable. But if that answer is 10% wrong, that doesn’t allow unalterable trust in the tool. In this it seems impossible to us that it can compete with Google today. On the other hand, it can show in an obvious way what search engines will look like in a few years.
Red Alert on Google
In any case, the release of this product caused a “red code” on Google, and the heads of the Mountain View company seem to have asked their engineers to speed up the development (which, of course, has been going on for many years) months, especially for LaMDA ) of equivalent technology. But it is clear that the dimensions of effective operation between ChatGPT and Google are not the same:
- the degree of error returned by ChatGPT today would not be accepted if Google had returned them to its official engine (but it might be the case with an experimental tool that will surely launch in 2023).
- the amount of information Brewed and indexed by Google (hundreds of billions of webpages) has no measurement in common with that of ChatGPT.
- The same goes for the volume of inquiries processed every second.
Sundar Pichai, CEO of Google, and Jeff Dean, Head of AI at Alphabet, state: This is an area where we must be bold and responsible. So we have to find a balance (…) For research, questions of veracity are really important; and for other applications, issues of bias, toxicity, and safety are also paramount “. Furthermore, Sam Altman, CEO of OpenAI, says no different and recently admitted the limitations of his tool: ” ChatGPT is incredibly limited, but good enough in some areas to give a false impression of being great. It would be a mistake to rely on it for anything important at this time (…) there is still work to be done in terms of robustness and truthfulness. »
So we certainly need to go beyond the “wow” aspect (which is clearly legit) from ChatGPT and think it’s one above all vision of what search engines will be in the medium term (Knowing that on the internet the notion of medium term is relative: we are talking about just a few years). It is clear that The “10 blue links” will soon disappear, at least in its current form. That’s also likely 2023 is a crucial year at that level and that Google will show us things soon, maybe as early as next May’s Google I/O event.
It remains to be seen how SEO will adapt to this new standard, these new standards of evolution. Because customization is probably important and not that easy…