[doe_eyes]

I think it's a pretty good and well-argued video, but how many folks watched it until the end, where the clip seamlessly pivots to plugging some commercial antivirus as a "cure" for the very problems it talks about?

There's so much irony in that... as with video monetization, Google sucks, except the alternatives don't work or suck worse. Since it's unlikely that Google will have a change of heart about product design, so the only way this can change is if we come up with a better way to search the internet.

[6510]

I walked away after 3 minutes or so and came back half an hour later. Apparently, (if one doesn't click skip) it takes half an hour to get 9 minutes into a 17 minute video. I put it back where I left off and got 2 more ads. Writing this I seek around the video a bit more and this some how provided an excuse to show even more ads. Then he ends with 70 seconds of smurfshark. Must be hard for the poor guy to make a living?

Ill say something interesting, since it seems so much needed atm.

When making things I never bother to look at the previous art the way one should. I love figuring things out by myself and usually end up with an inferior version of something fleshed out long ago. This is where the fun begins trying to beat that what exists. It hardly ever happens but I enjoy the process. [ill mark this as section A]

When trying to forge a word and string similarity test there were a lot of approaches to consider. They all worked but nothing is the level of correct that I was looking for. FOO is like FOE but FOO is also like FOOO. If I asked you to calculate how much FOO looks like FOE you could easily slap a number on it. Same for FOOO. It is just impossible to measure likeness as it is subjective but you cant really quantify it either. We could definitely say FOO is 66.666% like FOE and FOOO is 75% like FOO. But then FOOD is also 75% alike while we know FOOO is more like FOO than FOOD.

So I pick some of these "It's nonsense but it works!" and combine them into something that seems to work even better. I had successfully polished the turn it seemed.

I finally read some about popular algos trying to build a site search. One of the hacks was to ignore frequently occurring words. It was pretty funny as the data set I'm trying to search in and the key words I want to search are all frequently occurring. Like 90% of the set is black, 80% is large, 70% is heavy and I specifically need to be able to search "black large and heavy". Something like "to be or not to be" reducing to "".

What google does (and most of the industry) is value short articles with many relevant keywords over long[er] ones. Therefore Section A must be deleted. Much of what I wrote here must be deleted. I have to effectively stop being me and turn into a bland NPC in order to get noticed. Noticed for what exactly?

Maybe I'm a really annoying person unworthy of your attention? If one keeps polishing the turd in that direction eventually your attention will go to objectively annoying people who just want your time and your money.

Then, gradually, day by day, you spend enough time with these fckrs to start thinking it is normal and then, if you are not careful, you become one of them.