Honestly, bravo at BBC for whoever green-lighted this story. It puts the finger right on where it hurts for small publishers and illustrates Googles new way of thinking.
I am sad but also frightened for small creators and those who simply share interesting information once in a while. Google is bulldozing these people away from anywhere near the AI answers or the ten blue links below it.
The article didn’t mention it but people who got hit in September with the helpful update have still not recovered. Not a single site. This is the update that HouseFresh was hit with.
Small content creators are always going to experience randomness and chaos 'in the current system', cause the Amount of Content far exceeds the amount of Attention available.
If the platforms were honest they would signal to everyone how large that gap is. And people would make a more informed call about whether to get into content creation.
Instead of doing that the platforms just puts the whole onus on content creator "quality" - hey man if you aren't up in the ranking its your fault, maybe you are a looser, go do something more outrageous, scream a little louder, repeat yourself 24x7, spam everyone. Or buy our ads. Its a nice scam cause you get even large corps and presidents shelling out billions on ads.
It seems pretty clear that Google's era of search technology has aged out. SEO has been a game of cat and mouse since the web became an ad platform, it was only a matter of time before Google ran out of moves.
I'd much prefer a shift back to the earlier, smaller internet approach but it was driven by the enjoyment of exploration and discovery. Today most people prefer instant gratification online and just want answers and content thrown at them faster, with little regard for quality.
Google didn't get this far without providing much better search results than the competition for a while. I agree result quality has been declining for a while now, and the incentives of an ad and data business don't line up with providing quality search, but that hasn't always been the case.
Search before Google was much more rudimentary. Where they are today doesn't change that fact.
I don't think there's a new element to their thinking.
They used to rely on DMOZ's human curated content and then replaced it. Knowledge panels (at least the non-business ones) are reliant on wikipedia. The 10 blue links reliant on the web.
They don't mind borrowing until they can replace, and monetise what they borrowed.
This was inevitable. Traditional search is an inefficient way to access information in a world with large AI models. Google is being forced to do this due to competition from openAI, meta.ai, perplexity, etc.
LLMs cannot compete on truthful high-quality information from experts. Only on plausible babble. If that’s the only information you search for, then your perspective makes sense.
We shouldn’t be using LLMs for information retrieval at all. They should be summarizing individual web pages at best. That would make search more efficient. But to use plausible text generator without any guaranteed veracity checks in the pipeline for information retrieval is madness.
It is a bit surreal that you describe sucking up the world's information and spitting it out via a commercial entity and to you as a search result as 'inefficient'.
Perhaps it's more inefficient for all those content creators to bother creating anything vs what you actually derive from the search result, especially if they're not attributed/compensated.
Not necessarily. Internet search algorithms were designed to solve the discoverability problem. LLMs are generative, they're creating content rather than finding it for you.
They also serve a different business model. Sure you can have ads in LLM answers similar to Google search results, but Google is also selling clicks. LLMs are designed and used, at least today, to give you the answer so you don't need to see an original source.
Beyond that, the idea of a generative algorithm being used in tandem with sponsored results sounds pretty gross in my opinion. I already don't like Google's approach to search results, but if they're also generating the responses on the fly why am I even there? Its only a matter of time before the generative algorithm is trained specifically to drive more sponsorship revenue, maybe by making up content that works better for ad bids or stuffing a reference to a sponsor into a response only because the inline ad pays or converts better.
Seems like the option is constantly retrain an LLM or give it enough tools to interact with new data as needed. For the foreseeable future LLM's are going to need RAG or huge context windows which are effectively equivalent to RAG because you are isolating from some giant corpus your useful information you want the machine to meaningfully interact with.
It still seems like a different use case to me, but maybe I'm wrong there.
I don't mind a search algorithm including a summary as long as its true to the original, but I really don't want any generated content that's trying to predict how someone would answer my question. If LLMs end up replacing most of the use of tools like Google Search we really have moved past the discovery problem and don't necessarily want to find or see the original (human authored) content at all.
I am sad but also frightened for small creators and those who simply share interesting information once in a while. Google is bulldozing these people away from anywhere near the AI answers or the ten blue links below it.
The article didn’t mention it but people who got hit in September with the helpful update have still not recovered. Not a single site. This is the update that HouseFresh was hit with.
It’s insane.