“AI is going to fill the web with garbage. How are we going to deal with that? “
Is an excellent question. I genuinely worry about that. Several years ago, if you’re good at searching by keywords you can sift through low quality content and get to the reasonable sources easily. Then came the SEO spam. Blogs and articles that repeated content to appease the bots, but annoyed the users searching for content. Now, AI is going to put it on steroids. We are going to have LLMs search the web for us and eliminate results that look like they were generated by LLMs. What a pity state of the internet that will be.
Google is obviously run by the CFO nowadays, based on the company's behavior of the last few years. Google search quality is no longer a priority as long as ~90% of people are using it anyways since it's their devices default search engine and the requirements for search are basically "show me some news, a recipe, pictures of a travel location, where to find a product and perhaps a product review before I purchase".
Wait until these LLMs start eating the garbage they output; training themselves on their own hallucinations. I think soon people will be seeking out verified human-created, human-curated content.
That's already happening, and people already do. However, until it's proven that using LLMs to create content isn't categorically more profitable than using humans (which can be the case when the product is inferior, depending on the margins) then it isn't going to matter what people want. If 90% of the market is shit, people will learn to like the taste of shit.
I think a lot of people on this site seriously misunderstand the attention economy. Quality matters more and productivity matters less than people think.
ChatGPT type AI will (and does) happily make spelling and grammar mistakes if it makes the output more believable. It’s a believable-it’s-human output generator, not an actually-true output generator.
Historically, the two have been strongly correlated - but that connection has been weakening more and more over time, as the correlation is dependent on the actual knowledge level and ability to apply critical thinking to judge it.
It’s going to be the world’s best liar if it can be. Even if that means mostly telling the truth most of the time.
And considering ’organic’ certification allows a large whitelist of pesticides….
I don't know how Kagi does it, but it behaves exactly like old Google did before it got SEO-ed and now AI-ed to death. High quality search results, extremely user-oriented, and they treat every bug report seriously. It is absolutely fantastic and well worth the money.
he has nothing to say. his strategy has been to keep riding the gravy train for years. now the growth is slowing and he can see the core search business will be under attack and there is no answer. he'll essentially coast for the next 5 years. play politician, take few risks, "streamline operations" and fire a bunch of people, then walk off into the sunset. what is their future? cloud is still very small and has really difficult competition, youtube has challenging competition and isn't super profitable, maps/search/mail/docs/drive all feel quite stale. what's next?
> These days, questions for TGIF are pre-submitted and voted on. As Google has grown into a leaky behemoth, employees tell me these meetings have come to feel so scripted and benign that they no longer bother attending.
I enjoyed this excerpt from the article.
This is why I never attend all hands meetings anymore. Even at the peak of the layoffs and return to office turmoil, the all hands were held like nothing is happening. Employees show up to get answers, but instead it’s a panel about how awesome a particular leader is and their work on some nonsensical app.
I don’t miss out on anything by not going and get a few hours back every month. Win, win.
Working at Google wasn't a great experience (my time there was very short), but being able to ask Larry, Sergei and the rest of the SL team hard-hitting questions and get answers back was extremely cool. It's a shame that it turned into just another corporate all-hands.
26 comments
[ 0.19 ms ] story [ 69.3 ms ] threadFor this humans should be willing to pay another human for their efforts. Which history has shown would be tough.
And regardless, which have real content when the reviews and ads are also generated?
As human are fallible, a mistake in a book would show that it's "human made".
Or start a "no ai used" certification process like for organic vegetables?
Historically, the two have been strongly correlated - but that connection has been weakening more and more over time, as the correlation is dependent on the actual knowledge level and ability to apply critical thinking to judge it.
It’s going to be the world’s best liar if it can be. Even if that means mostly telling the truth most of the time.
And considering ’organic’ certification allows a large whitelist of pesticides….
I enjoyed this excerpt from the article.
This is why I never attend all hands meetings anymore. Even at the peak of the layoffs and return to office turmoil, the all hands were held like nothing is happening. Employees show up to get answers, but instead it’s a panel about how awesome a particular leader is and their work on some nonsensical app.
I don’t miss out on anything by not going and get a few hours back every month. Win, win.