Where Algorithmic Search and Generative AI part company

... and where the latter actively antagonises the former

[ronin]

Elsewhere in Foo, I suggested that Generative AI like ChatGPT may come to represent a third type of web-based oracle, alongside algorithmic search engines (like Google) and crowdsourced editorial (like Wikipedia).

My thought was that Generative AI would end up adopting a position complementary to Algorithmic Search.

This Wired article by Will Knight focuses on ways in which both forms of information enquiry are entirely unlike each other and how the upstart may even turn out to be downright antagonistic towards its more established counterpart:

The Race to Build a ChatGPT-Powered Search Engine
[wired.com...]

The efforts of search giants notwithstanding:

Microsoft, which has invested around $10 billion in ChatGPT’s creator OpenAI, is said to be somehow adding the underlying technology to its second-ranked search engine Bing. [...] Google [...] is reported to be scrambling to respond. [...] China’s leading search engine, Baidu, is working on a Chinese language bot similar to ChatGPT.

this is very much worth considering:

the way the technology works is in some ways fundamentally at odds with the idea of a search engine that reliably retrieves information found online. There’s plenty of inaccurate information on the web already, but ChatGPT readily generates fresh falsehoods. Its underlying algorithms don’t draw directly from a database of facts or links but instead generate strings of words aimed to statistically resemble those seen in its training data, without regard for the truth.

In a worst-case scenario, the emergence of generative AI threatens to drown the algorithmic search that we all know and love (and which is arguably, less competent now in 2023 than it was, 16-17 years ago) in a flood of impenetrably convincing noise:

Gary Marcus, a professor emeritus at New York University and a vocal critic of AI hype, believes ChatGPT is unsuited to search because it has no true understanding of what it says. He adds that tools like ChatGPT may cause other problems for search companies by flooding the internet with AI-generated, search engine-optimized text. “All search engines are about to have a problem,” he says.

[Sgt_Kickaxe]

There is a divergence between people who care and people more focused on what it means to their bottom line. More are joining the later group every month.

When search traffic is required and the company providing it makes tens of billions per quarter, and many who need that traffic live like they are step away from needing a soup kitchen, people stop caring about search "innovation" and want things back the way they were.

Can't blame them.

[ronin]

This article in TIME by Andrew R. Chow - which provides an excellently-summarised timeline of the emergence and growth of Generative AI from the mid-2010s onwards - broadly agrees that Generative AI may not be a very good complement for Algorithmic Organic Search at all:

The AI Arms Race Is On. Start Worrying
[time.com...]

Alphabet and Microsoft are most interested in how AI will make their search engines more valuable, and have shown demos of Google and Bing in which the first results users see are AI-created. But Margaret Mitchell, chief ethics scientist at the AI-development platform Hugging Face, argues that search engines are the “absolute worst way” to use generative AI, because it gets things wrong so often. Mitchell says the actual strengths of AIs like ChatGPT—assisting with creativity, ideation, and menial tasks—are being sidelined in favor of shoehorning the technology into moneymaking machines for tech giants.

Google already prioritizes paid ads in search results. It’s not much of a leap to imagine it doing the same with AI-generated results. [But] If humans come to rely on AIs for information, it will be increasingly difficult to tell what is factual, what is an ad, and what is completely made up.

[not2easy]

The Time article is excellent as far as I got, but I had to duck out as more ad and email popups cover the text and scrolled me past what I was reading. I think I got 2/3rds of it anyway. It does explain well why ChatGPT and other AI generated content is a bad idea the way that search engines are using it though. It is as likely to output a nicely crafted piece of fiction as a useful response. It is more likely to feed alternative realities based on how a query is entered. Yes, they need to apply the brakes.