5 ms·

I once switched to Google, one day in '98, when they were basically ad-free, with good results. I once switched away from Google, one day in '23, when the firs

I once switched to Google, one day in '98, when they were basically ad-free, with good results. I once switched away from Google, one day in '23, when the first seven results for many queries were 'sponsored', and shitty. These days, I let LLMs answer most of my questions. They are sufficiently easy to use. For regular surfing, I rely on Kagi. Both services are pay for play, but my lifetime is not unlimited, and can be spent better than shifting through advertisements.

LeoPanthera 2y ago
You've got to be careful about which LLM. I asked Microsoft Copilot for advice about staying in Paris, and then as a followup in the same thread, asked about navigation apps. It started recommending apps that work well... in New York City. If Copilot can't retain context, in the same thread, after a single message, then it's useless. ChatGPT, which apparently Copilot is based on, didn't have this problem. And from the linked article I have just discovered "Perplexity", which looks a bit like ChatGPT only plugged into a search engine. ChatGPT is also plugged into a search engine, so maybe it's plugged in... more? Not sure. If anyone working at Perplexity is reading this, please add an "about" page to tell me what the site actually does. Nowhere I could find would explain. Everything assumes you already know, and I eventually had to go to Wikipedia to find out. If anyone makes an LLM which only knows what's in Wikipedia, and doesn't hallucinate anything else, they would clean up.
- smackeyacky 2y ago
  Why not just use the search bar in Wikipedia and call it good?
  - LeoPanthera 2y ago
    Because "searching for an encyclopedia article" and "asking questions about a branch of knowledge" are fundamentally different approaches. In theory, an LLM that was a talking Wikipedia would not be limited to giving you knowledge from just one page, it could extract multiple pieces of information from multiple articles to answer the precise question you had, and could accept follow-up questions on the same or similar topics.
    - card_zero 2y ago
      It's a good idea, but it might turn out to hallucinate even more. For instance, there is one article on Abraham Lincoln and one on Abraham Lincoln: Vampire Hunter, so the statistics (50/50) in regard of mixing fantasy with reality are perhaps discouraging.
- readyman 2y ago
  [dead]
- wslh 2y ago
  If you know about the subject LLMs makes you turbo. Even when you don't know about the subject a rational thinking background gives you more legs. LLMs are incredible tools, it is expected that they are not the only ones in your toolbox. When I take a random photo and asks ChatGPT to describe it I always wonder how I saved a whole lifetime (or many ones) dedicated to computer vision research and can build upon all of this.
- Zambyte 2y ago
  > ChatGPT is also plugged into a search engine, so maybe it's plugged in... more? Not sure. I haven't used ChatGPT search, but I have used Perplexity. The way that it works, it will potentially do like a dozen searches to try to understand your question (if it needs to), search individual topics in the question, search joint topics in the question, and then take 20 results or so, including web pages and videos, and synthesize an answer to your question. > If anyone makes an LLM which only knows what's in Wikipedia, and doesn't hallucinate anything else, they would clean up. Kagi has a "lens" feature that lets you scope down web results to certain sources. It's supposed to be able to be used in their AI assistant, but it's currently ignored / broken there. They are working on a large overhaul (I think it's supposed to be out by the end of this month?), hopefully that works after that. You could achieve what you described using that then.
stordoff 2y ago
LLMs are also capable of mixing up information about different people: > Prompt: Who is Tom Faber > ChatGPT: Tom Faber is a versatile individual with multiple professional identities. He is known as a culture journalist and gaming reviewer, contributing to prestigious publications like the Financial Times, The Guardian, and The New York Times. His work often covers topics related to arts and gaming. Additionally, Tom Faber serves as the Clinical Director of the Zuni Indian Health Service Hospital and is a founding member of the Zuni Youth Enrichment Project (ZYEP), where he has been instrumental in developing programs for the Zuni community since 2006 (Zuni Enrichment) . It's probably obvious here that it's talking about two different people as the fields are unconnected, but if they were closer, you'd assume it's the same person. They can also invent things: > Prompt (new chat): tell me about tom faber > Tom Faber is a prominent figure in British literature, known for his work as an author and editor. His most notable works include his novel "The Scent of Death," which delves into themes of mystery and psychological drama. Faber's writing often features intricate plots and rich character development, reflecting his deep understanding of human nature and storytelling. The only book titled "The Scent of Death" I can find was written by Simon Beckett.
bigstrat2003 2y ago
> These days, I let LLMs answer most of my questions. That is a great route to being duped by LLM hallucinations, but not so much for actual knowledge.

LeoPanthera 2y ago

You've got to be careful about which LLM. I asked Microsoft Copilot for advice about staying in Paris, and then as a followup in the same thread, asked about navigation apps. It started recommending apps that work well... in New York City. If Copilot can't retain context, in the same thread, after a single message, then it's useless. ChatGPT, which apparently Copilot is based on, didn't have this problem. And from the linked article I have just discovered "Perplexity", which looks a bit like ChatGPT only plugged into a search engine. ChatGPT is also plugged into a search engine, so maybe it's plugged in... more? Not sure. If anyone working at Perplexity is reading this, please add an "about" page to tell me what the site actually does. Nowhere I could find would explain. Everything assumes you already know, and I eventually had to go to Wikipedia to find out. If anyone makes an LLM which only knows what's in Wikipedia, and doesn't hallucinate anything else, they would clean up.

smackeyacky 2y ago
Why not just use the search bar in Wikipedia and call it good?
- LeoPanthera 2y ago
  Because "searching for an encyclopedia article" and "asking questions about a branch of knowledge" are fundamentally different approaches. In theory, an LLM that was a talking Wikipedia would not be limited to giving you knowledge from just one page, it could extract multiple pieces of information from multiple articles to answer the precise question you had, and could accept follow-up questions on the same or similar topics.
  - card_zero 2y ago
    It's a good idea, but it might turn out to hallucinate even more. For instance, there is one article on Abraham Lincoln and one on Abraham Lincoln: Vampire Hunter, so the statistics (50/50) in regard of mixing fantasy with reality are perhaps discouraging.
readyman 2y ago
[dead]
wslh 2y ago
If you know about the subject LLMs makes you turbo. Even when you don't know about the subject a rational thinking background gives you more legs. LLMs are incredible tools, it is expected that they are not the only ones in your toolbox. When I take a random photo and asks ChatGPT to describe it I always wonder how I saved a whole lifetime (or many ones) dedicated to computer vision research and can build upon all of this.
Zambyte 2y ago
> ChatGPT is also plugged into a search engine, so maybe it's plugged in... more? Not sure. I haven't used ChatGPT search, but I have used Perplexity. The way that it works, it will potentially do like a dozen searches to try to understand your question (if it needs to), search individual topics in the question, search joint topics in the question, and then take 20 results or so, including web pages and videos, and synthesize an answer to your question. > If anyone makes an LLM which only knows what's in Wikipedia, and doesn't hallucinate anything else, they would clean up. Kagi has a "lens" feature that lets you scope down web results to certain sources. It's supposed to be able to be used in their AI assistant, but it's currently ignored / broken there. They are working on a large overhaul (I think it's supposed to be out by the end of this month?), hopefully that works after that. You could achieve what you described using that then.

stordoff 2y ago

LLMs are also capable of mixing up information about different people: > Prompt: Who is Tom Faber > ChatGPT: Tom Faber is a versatile individual with multiple professional identities. He is known as a culture journalist and gaming reviewer, contributing to prestigious publications like the Financial Times, The Guardian, and The New York Times. His work often covers topics related to arts and gaming. Additionally, Tom Faber serves as the Clinical Director of the Zuni Indian Health Service Hospital and is a founding member of the Zuni Youth Enrichment Project (ZYEP), where he has been instrumental in developing programs for the Zuni community since 2006 (Zuni Enrichment) . It's probably obvious here that it's talking about two different people as the fields are unconnected, but if they were closer, you'd assume it's the same person. They can also invent things: > Prompt (new chat): tell me about tom faber > Tom Faber is a prominent figure in British literature, known for his work as an author and editor. His most notable works include his novel "The Scent of Death," which delves into themes of mystery and psychological drama. Faber's writing often features intricate plots and rich character development, reflecting his deep understanding of human nature and storytelling. The only book titled "The Scent of Death" I can find was written by Simon Beckett.

bigstrat2003 2y ago

> These days, I let LLMs answer most of my questions. That is a great route to being duped by LLM hallucinations, but not so much for actual knowledge.