Comment by thefourthchime | Hacker News Reader

Comment by thefourthchime | Hacker News Reader

For those who are lazy like me—or maybe you'd call it efficient—here's the key points summary from Kagi:

- Google Search has evolved significantly from its early days as a simple list of blue links to becoming an encyclopedia, predictive engine, image repository, shopping mall and more that is overloaded with information.

- It has become more difficult to find authoritative answers on Google Search due to an overabundance of sponsored content, prompts, and low-quality keyword-stuffed pages.

- Google is currently undergoing an antitrust trial regarding whether it maintains its search engine monopoly through anticompetitive means such as exclusivity deals rather than having truly superior technology.

- Access to large amounts of user data is very important for powering search engine algorithms through personalization and improving the user experience.

- While Google argues diminishing returns to scale for user data, internal emails show Google engineers acknowledging that scale remains highly important for the company.

- Some feel Google has lost its way and become conservative due to its success, with its corporate bureaucracy stifling acquired companies.

- Google Search has become bloated with advertisements and prioritization of its own services over organic results.

- Search quality may be declining as its algorithms are gamed by low-quality sites and search engine optimization techniques.

- Google's trajectory of scaling up its mission to organize the world's information through acquisitions and exclusivity deals has put pressure on it to keep growing.

- No company can likely grow indefinitely to keep up with the ever-expanding internet without being overwhelmed, as Google Search now demonstrates.

Thanks for the excellent summary. Though I think the article and summary miss the root cause: Google Search has two masters, users and advertisers. These stakeholders’ interests are not aligned, and by taking one middle road after another, Google has deteriorated into near-uselessness for end users, relying on scale rather than quality.

> Google Search has two masters, users and advertisers

No, it doesn't. It has one master: advertisers. Advertisers are the customer. Users are the product. Search results are the bait.

This is self-evidently true. Your customer is the one who pays for your services. Google's users don't pay, never have, never will, because search is not a product you can sell.

User attention, on the other hand, is a very valuable commodity for which people will pay handsomely.

> > Google Search has two masters, users and advertisers

> No, it doesn't. It has one master: advertisers. Advertisers are the customer. Users are the product. Search results are the bait.

I wouldn't put it that strictly. Users are still masters in that they can continue to use Google or not.

You’re engaging in some sleight of hand by first saying attention is not payment, and later that it is.

The currencies are different but Google absolutely needs to satisfice two disparate customers.

No. Users are engaging in a barter transaction. It's not different currencies, it's the complete absence of currency. Barter transactions are symmetric. You could as easily say that Google is the user's customer.

Users are Google customers in the same way that (say) fish are a fisherman's customer. Fisherman engage in barter transaction with fish, offering them bait (food, or at least the prospect of food) in exchange for allowing themselves to be sold as a product to a third party who gives the fisherman actual money. That is what makes the third party the fisherman's customer and not the fish. Likewise, Google engages in barter transactions with users, offering them bait (useful information, or at least the prospect of useful information) in exchange for allowing their attention to be sold to a third party for actual money. The money is what makes the advertiser Google's actual customer and not the user.

Apologies, thought we were talking economics, where barter is every bit as legitimate as cash. But you’re talking polemics where there’s moral failing in Google’s and user’s parts. Fair enough.

I didn't say barter was illegitimate, I said it was symmetric. There are salient differences between barter transactions and currency transactions even from a purely economic perspective. It has nothing to do with morality or polemics. There is a reason money is a thing.

> Google is currently undergoing an antitrust trial regarding whether it maintains its search engine monopoly through anticompetitive means such as exclusivity deals rather than having truly superior technology.

I'm not really sure how they lose that case -- Firefox switched to Yahoo after MS outbid them, and switched back to Google the second they could legally do so and keep the money.

I'm still not sold on the idea that "personalization" improves the "user experience" of search.

When I think about what I actually use search for... it's highly situational. It depends as much on day-to-day and moment-to-moment context as it does any long-term trend specific to myself.

Whatever "personalization" could be realized over search results, would it really offer that much benefit over improving the search function to sift through and find results that are just more useful generally?

What I mean is there seems to far, far more to be gained from making search results better fit the query than to fit me personally.

That’s a pretty great summary. Can you elaborate a bit on how this works with Kagi? Is it just a wrapper over gpt-4, do you feed it an article and ask for a summary, or does it summarize in the search results?

Just go here

https://kagi.com/summarizer/

But the article has a paywall, can kagi summarizer bypass the paywall itself?

When I was using Bing AI or Google Bard to summarize an article, they often refuse to do so because of paywall or some other reasons.

Good question. I've been using Kagi a long time but have never tried this feature. I just tried, and it does bypass the paywall. No idea which sites that works on, but I'd imagine it correlates with ons archive.is can bypass - I assume crawlers are just allowed to read the whole thing.

Why is it that when I use Google search, the links are always some intermediary google.com/** links which when clicked takes me to the original site. Is this some kind of user tracking?

Yes

I would add, as advertiser this time, that you don't know what you are really paying in AdWords. I used AdWords from the beginning and there were business leads for super niche spaces, now it is very difficult to get leads with a moderated budget and you cannot explain clearly why.

Apparently (since this is the top-ranked comment) I'm alone in thinking that this sort of "summary" is dumb and pointless. Just skim over the article. You cannot separate form and content. What's next -- a PowerPoint summary?

Reply on news.ycombinator.com