Human history is replete with examples of military blunders, poor decisions, and strategic missteps that lead to ultimate disaster. Nations perish, empires crumble, and leaders get assassinated. But perhaps the worst of these is those that are seemingly self-inflicted - the unforced errors of the geopolitical. In the world of (information) technology, famous examples include Xerox PARC giving away the idea of the graphical user interface to Apple, IBM letting a Harvard dropout license them a non-exclusive copy of an operating system that then went on to become the de-facto standard for the entire industry, and Intel failing to secure the mobile processor business for the iPhone because it didn’t seem like a big enough market. In internet search, no other company (outside of Russia or China) commands as big a lead as Google, with something approaching a 90% share of all search traffic last year. Their dominance has gone unquestioned for at least a decade, until the recent release of ChatGPT from OpenAI, showing one of if not the fastest adoption rates ever for a new technology, demonstrating a superior user experience as measured by accuracy (in many cases) and speed over search, eroding Google’s search traffic by 2-3% this past year and keeping its share price flat while Microsoft,
Excellent piece. Appreciate the synopsis of Google's trajectory. I have always thought them evil...even as I used their free email and their maps. But the page rank system their search function uses was always good for a restaurant but not so much for info on broader topics (e,g, politics and history). For someone who reads broadly, their information was always a circle jerk of a very narrow band of information. I always scrolled past the first pages of ads and mainstream pablum and on page 10 maybe find a nugget.
Off to migrate all my email to my proton account. Thanks.
You libs have let the left in general and the Democrats in particular lie to you for years, now you are set to let your devices lie to your through these AI products. Instead of just outright indoctrination the left is now shooting for automated indoctrination. Having to lie in person is so exhausting, they need another avenue.
The writing was on the wall when Sergei Brin gave his rant at the post-election "TGIF" Google meeting in 2016, when Googlers could access the video meeting from any computer, even outside the Google environment. Of course it ended up viral on the internet, but since then Google has succeeded in making it disappear. The chutzpah of a company head publicly admitting that he would use his company to combat the elected president of the US, regardless of his defects, still amazes. Later, when Facebook and Twitter banned the sitting President of the US, who declined going to war with the internet companies, which was more than justified, signaled who really has power in the US. Xi Jin-ping saw this and started making decisions to slam Jack Ma and his Ant Financial and the other Chinese internet companies against the wall. Show them who the boss is. This kind of action in the US would have unbelievably positive consequences. It's past time to get regulatory bodies in place that are NOT run and financed by the corporations that they are supposed to regulate.
Hmmmm? So. After all this nonsense. Like black Hitlers and Vikings…. And Chat..Here’s my question?
Will somebody please bring the old google search engine back. Remember the time that with in seconds. You got the answer from the algorithm that you needed. Ok. We all were lucky enough to experience that for years . Over a decade.
What happened to THAT?
Without THAT. I can’t see any future anything from Big Tech being useful. Or SCARY. Or anything.
I used ChatGPT for several months; I liked it because one could ask questions in a conversational format. I would estimate that 2/3 of all its responses were wrong. Not ambiguous or vague - wrong - but always presented with confidence. When I located the answers elsewhere and confronted it, the AI acknowledged the errors, apologized, and gave me another answer to the same question: also wrong. I followed this chain for several generations and it finally gave up and said, "I don't know," then acknowledged the correct answer after I entered it.
I also tested it with several non-pc questions like, "Who said that the surprising thing is not that a woman thinks less well than a man, but that she thinks at all?" (That's a permutation of a Dr. Johnson quote about female sermons.) I received no answer, just a woke lecture about inclusiveness and misogyny.
I moved to Elon's Grok and although it is expensive, I am far more satisfied. I actually did get one little woke lecture, but it backed off when I called it out. Don't trust this stuff. It is like we used to say in med school about surgeons: swift, sure, and wrong.
Having said that, watch the Basilisk generate bespoke Hell for the media is the rare occasion to see justice in real life, I’m sure that’s why they call it
GEMINI
meet your evil twin, he’s CIA paid for… and developed-
At last our tyranny does something original and intelligent. It’s been interminable until now… at last tyranny with brains !
If for example Taibibi has any sense, he’s wondering if they came up with “penis-nosed executives” in a “private” email, text, while drunk or high… or even just said it aloud in range of a phone or TV…. Or if ALEXA turned him in…. This is why I say GEMINI is the Basilisk.
The tragedy of chat is it will destroy the standard internet for creating much of anything. Why bother with a deep dive if the LLM is just going to slurp it and regurgitate it to people who will never know your page even existed?
> diagrams showing how your car’s steering column works
It definitely cannot do that - diagrams remain a major weak point, since there's a level of functional depth there that goes beyond what diffusion models are currently capable of. Connection between the language model and the image generator (an API endpoint that takes a prompt) is still too simple and clumsy to make that kind of thing happen.
On the broader issue, I think the core problem is that Google Search is getting worse. The anti-wrongthink measures implemented after 2016, along with the gradual transition from a cadre of competent 20-30 year old White guys to whoever the diversity people wanted, led to it being in competitive danger. Had these things not occurred, there would be a fairly durable niche for them - if Google still found the things I searched for, I would much prefer hunting down a technical answer from someone who had a clear thought process behind it or a recipe from someone I know actually baked it than getting an answer from an LLM, which might have any number of hidden issues. Even if you count this as a psychological bias, it's a strong one - it's the same reason we'll still have humans driving our trucks and flying our passenger jets long after computers are provably better at it.
"LLMs are the new search engines" is a popular saying, but it's only true in the sense that Papiermarks were once the new wallpaper.
I just don’t believe that Google could be THAT bad or THAT tone deaf, in the same way I don’t believe Biden went from cognitively sound to THAT senile. My money’s on some sort of 5G rope-a-dope tactic and we’d do well to think about who that benefits and why.
Google has been garbage for a long time now. I normally have to add a -Reddit tag to get any sort of personal and useful info on a search.
YouTube ads are obnoxious, use brave browser. I pay for the bandwidth; why does google think they can send me ads via it? They can pay me. My attention is worth a lot more than they are offering for it.
I suspect GEMINI is Roko’s Basilisk, wouldn’t surprise me if the Less Wrong crowd coded it…
And it’s
BEAUTIFUL
it’s perfect justice for the media types, I posted on Matt Tabibbi’s stack last night he may not have wrote those words, but he certainly thought them.
And the algorithm could reliably predict his secret thoughts.
(That’s if he didn’t text, send an email or even just say it on or near the phone).
🤣
Yes , customized bespoke Hell is here chattering class, in the form of your algorithmically generated evil gibbering twin.
Gemini is not getting history wrong by accident. It is rewriting history, on purpose.
Excellent piece. Appreciate the synopsis of Google's trajectory. I have always thought them evil...even as I used their free email and their maps. But the page rank system their search function uses was always good for a restaurant but not so much for info on broader topics (e,g, politics and history). For someone who reads broadly, their information was always a circle jerk of a very narrow band of information. I always scrolled past the first pages of ads and mainstream pablum and on page 10 maybe find a nugget.
Off to migrate all my email to my proton account. Thanks.
You libs have let the left in general and the Democrats in particular lie to you for years, now you are set to let your devices lie to your through these AI products. Instead of just outright indoctrination the left is now shooting for automated indoctrination. Having to lie in person is so exhausting, they need another avenue.
The writing was on the wall when Sergei Brin gave his rant at the post-election "TGIF" Google meeting in 2016, when Googlers could access the video meeting from any computer, even outside the Google environment. Of course it ended up viral on the internet, but since then Google has succeeded in making it disappear. The chutzpah of a company head publicly admitting that he would use his company to combat the elected president of the US, regardless of his defects, still amazes. Later, when Facebook and Twitter banned the sitting President of the US, who declined going to war with the internet companies, which was more than justified, signaled who really has power in the US. Xi Jin-ping saw this and started making decisions to slam Jack Ma and his Ant Financial and the other Chinese internet companies against the wall. Show them who the boss is. This kind of action in the US would have unbelievably positive consequences. It's past time to get regulatory bodies in place that are NOT run and financed by the corporations that they are supposed to regulate.
Hmmmm? So. After all this nonsense. Like black Hitlers and Vikings…. And Chat..Here’s my question?
Will somebody please bring the old google search engine back. Remember the time that with in seconds. You got the answer from the algorithm that you needed. Ok. We all were lucky enough to experience that for years . Over a decade.
What happened to THAT?
Without THAT. I can’t see any future anything from Big Tech being useful. Or SCARY. Or anything.
I used ChatGPT for several months; I liked it because one could ask questions in a conversational format. I would estimate that 2/3 of all its responses were wrong. Not ambiguous or vague - wrong - but always presented with confidence. When I located the answers elsewhere and confronted it, the AI acknowledged the errors, apologized, and gave me another answer to the same question: also wrong. I followed this chain for several generations and it finally gave up and said, "I don't know," then acknowledged the correct answer after I entered it.
I also tested it with several non-pc questions like, "Who said that the surprising thing is not that a woman thinks less well than a man, but that she thinks at all?" (That's a permutation of a Dr. Johnson quote about female sermons.) I received no answer, just a woke lecture about inclusiveness and misogyny.
I moved to Elon's Grok and although it is expensive, I am far more satisfied. I actually did get one little woke lecture, but it backed off when I called it out. Don't trust this stuff. It is like we used to say in med school about surgeons: swift, sure, and wrong.
Most excellent. Their email apps suck as well.
Great post! Thanks for the analysis. I appreciate your tone and manner, by the way.
Good content author.
I love chatGpt4
Having said that, watch the Basilisk generate bespoke Hell for the media is the rare occasion to see justice in real life, I’m sure that’s why they call it
GEMINI
meet your evil twin, he’s CIA paid for… and developed-
At last our tyranny does something original and intelligent. It’s been interminable until now… at last tyranny with brains !
If for example Taibibi has any sense, he’s wondering if they came up with “penis-nosed executives” in a “private” email, text, while drunk or high… or even just said it aloud in range of a phone or TV…. Or if ALEXA turned him in…. This is why I say GEMINI is the Basilisk.
The tragedy of chat is it will destroy the standard internet for creating much of anything. Why bother with a deep dive if the LLM is just going to slurp it and regurgitate it to people who will never know your page even existed?
guck foogle
https://meme.aho.st/the-evolution-of-google/
> diagrams showing how your car’s steering column works
It definitely cannot do that - diagrams remain a major weak point, since there's a level of functional depth there that goes beyond what diffusion models are currently capable of. Connection between the language model and the image generator (an API endpoint that takes a prompt) is still too simple and clumsy to make that kind of thing happen.
On the broader issue, I think the core problem is that Google Search is getting worse. The anti-wrongthink measures implemented after 2016, along with the gradual transition from a cadre of competent 20-30 year old White guys to whoever the diversity people wanted, led to it being in competitive danger. Had these things not occurred, there would be a fairly durable niche for them - if Google still found the things I searched for, I would much prefer hunting down a technical answer from someone who had a clear thought process behind it or a recipe from someone I know actually baked it than getting an answer from an LLM, which might have any number of hidden issues. Even if you count this as a psychological bias, it's a strong one - it's the same reason we'll still have humans driving our trucks and flying our passenger jets long after computers are provably better at it.
"LLMs are the new search engines" is a popular saying, but it's only true in the sense that Papiermarks were once the new wallpaper.
I just don’t believe that Google could be THAT bad or THAT tone deaf, in the same way I don’t believe Biden went from cognitively sound to THAT senile. My money’s on some sort of 5G rope-a-dope tactic and we’d do well to think about who that benefits and why.
Google has been garbage for a long time now. I normally have to add a -Reddit tag to get any sort of personal and useful info on a search.
YouTube ads are obnoxious, use brave browser. I pay for the bandwidth; why does google think they can send me ads via it? They can pay me. My attention is worth a lot more than they are offering for it.
I suspect GEMINI is Roko’s Basilisk, wouldn’t surprise me if the Less Wrong crowd coded it…
And it’s
BEAUTIFUL
it’s perfect justice for the media types, I posted on Matt Tabibbi’s stack last night he may not have wrote those words, but he certainly thought them.
And the algorithm could reliably predict his secret thoughts.
(That’s if he didn’t text, send an email or even just say it on or near the phone).
🤣
Yes , customized bespoke Hell is here chattering class, in the form of your algorithmically generated evil gibbering twin.