The author’s conclusion has me wondering what would it take to build a really good new system? Could we make a paid version, at a cost users would find reasonable to grow to a large enough base, and one that is incentivized to find users the best links quickly? Unlike how Google has been moving in recent years
As mentioned in the article, what it would take is a minimum of a billion dollars a year.
There isn’t a lot of market for a paid search engine. Kagi are trying, and have about 30,000 subscribers, but that’s a tiny drop in the bucket. If all those subscribers were on the unlimited plan ($10 a month), then they are bringing in 3.6 million in revenue a year.
If the search index itself costs a billion dollars to maintain, they also need to cover payment fees, support costs, and of course the cost of building and maintaining a website to actually access that search index.
If we wanted to fund it on donations, we need at least 1,000,000 people to donate $1,000 each per year.
Google had a great environment to start their search engine because the internet was small. Now you have an internet that’s more than half bot traffic and billions of websites generated by AI, on top of all the good stuff. Plus the expectation of real time search results, that Google didn’t have to deal with before web 2.0.
I am not convinced a new traditional search engine can compete. Any true competitor or successor to Google will probably have to do what they did: completely redefine search.
This is Lemmy so I’m gonna hit the screw with a hammer and suggest maybe some sort of federated search with an instance trust model that lets each instance take care of only a small part of the web but have a way of boosting results from more trustworthy instances.
Add comment