AI chatbots are not a replacement for search engines
iai.tv
iai.tv
- Today's search engines will give you links to millions of documents
- ChatGPT, if asked correctly, will instead generate one perfect document based on millions of the documents
To me, that's a clear evolution of the search engine, especially with all the SEO & ad spam that's plaguing Google and others currently.
I wasn't ready to pay a monthly fee for an ad-free Google. I am ready to pay for something like ChatGPT.
Google has an issue on their hands and is probably working overtime to lobby the threat of ChatGPT away.
"What's a good social website for people of a technical bent"
It recommended: Reddit, Hacker News, Stack Exchange, Quora, LinkedIn, ResearchGate, GitHub
That's a pretty good list!
Tried the same search on Google, and the first result was some spammy site "15 top social networks," #1 being facebook. Not very useful.
I definitely think chatbots will replace many uses of search engines. It's already way more useful for lots of stuff for me.
The biggest advantage Google currently has is their massive indexing ability. ChatGPT is oblivious to something that just happened (ie was posted) an hour ago, a day ago, heck, a month ago.
> Without more information, I cannot estimate the accuracy or reliability of these claims. It is important to note that there may be conflicting or inaccurate information available online, and it is always a good idea to verify information from multiple sources before accepting it as fact.
which seems pretty reasonable.
There are many self-hosted solutions that could potentially meet the requirements you have listed for your RSS aggregator and reader project. Here is a list of some popular options, ranked in order of their approximate percentage of compatibility with the features you have specified:
RSS Aggregator
| Percentage Score | URL |
| --- | --- |
| 90% | https://tt-rss.org/ |
| 85% | https://miniflux.app/ |
| 80% | https://newsboat.org/ |
| 75% | https://github.com/lobsters/lobsters |
| 70% | https://github.com/noahcoad/feeds |
| 65% | https://github.com/ssut/py-googlenews |
| 60% | https://git.gnome.org/browse/feedreader/ |
RSS Reader
| Percentage Score | URL |
| --- | --- |
| 90% | https://tt-rss.org/ |
| 85% | https://miniflux.app/ |
| 80% | https://newsboat.org/ |
| 75% | https://github.com/lobsters/lobsters |
| 70% | https://github.com/noahcoad/feeds |
| 65% | https://github.com/ssut/py-googlenews |
| 60% | https://git.gnome.org/browse/feedreader/ |
Please note that these percentages are approximate and are intended only as a rough guide. You may want to consider other factors such as the user interface, performance, security, and integration with other tools when choosing a solution for your project.
Oh it tried to convince me that four is not larger than one. :) It went like this:
At first I asked ChatGPT to summarise J.C. Owen’s contribution to geometric constraint solving. It gave me an answer which might be even correct. It sounded correct anyway. Then I have asked for a worked example of his method used on a toy geometric constraint system. It run into problems with that. Chiefly that the system is selected was under constrained.
When I asked about that it insisted that it is well constrained. It even volunteared the info that for a system to be well constrained it needs as many variables as constraint equations. Then it told me that the system it generated has 4 variables and 1 equation.
As a last question I asked if “Is four larger than 1?” and it resolved the seeming contradiction by telling me that in the context of geometric constraint solving four is not larger than one. One does learn a new thing every day. :)
Same error a human might make if they read too many papers too quickly. Or went to a film festival and was asked to describe the plot(s) of the movies they watched.
Calvin:how do they know the weight limit for a bridge?
Dad: they drive bigger and bigger trucks over it until it collapses, then they rebuild the bridge and put up the sign
https://www.reddit.com/r/atheism/comments/70mdfj/recently_go...
At no point was it trying to convince you of something.
At no point does it have any clue that "events occur in some order". Or that "time is ordered"...
At no point will you EVER be able to "convince it otherwise".
This is a language model. All it knows is that "in my training set, something said something akin to the two English words: 'ordered events'". It has no model of facts, deduction, or logic, as long as its results are a reasonable approximation of the English language as influenced by your prompt.
The lack of self-reflection, of logical reasoning ability and associated memory is what makes ChatGPT useless for the use cases I’d actually find interesting.
I think we are forgetting that Google suffers from these problems too. A human who asks bad questions will continue to ask bad questions. AI should in general be trained to do what the human wants. If the human wants something stupid, that's on the human. Google also isn't guaranteed to find the right information you're looking for, especially if it is on an obscure topic or if the query is highly overloaded or being taken over by current trends.
GPT vomits up word soup. How should you expect to verify every factoid contained? As others have pointed out: if you ask GPT to cite its sources it will very happily, and authoritatively, spit back articles and authors that do not exist.
I have asked chatGPT about my boss (a well published researcher) and it managed to summarise his work quite well, name the field he has worked in and even write about some of his previous projects. It also insisted on that he is already dead. Which made it very funny when I have shown it to him. :)
But funny doesn’t win one a trophy in information retrieval. Clearly it knows a lot about a lot of things, but the accuracy is hit and miss. Can this be fixed? Then this is the future. If it can’t, because this is a fundamental property of these systems then it won’t be a usefull replacement of search.
People are okay-ish with Google in part because it drives traffic and traffic can be profitable. Fancy features that snatch website content and show them in Google's result pages are already not appreciated by indexed websites.
If generative search becomes the dominant interface, we will eventually see severe public info stagnation until alternative business models can grow around it or avoid it altogether. I suspect we'll see more platforms like Spotify for X and a continued shift toward subscription platforms and youtube.
Of course, what's to stop the bots from watching all of Netflix, listening to every podcast, etc? It will be an interesting decade for law / regulation / licensing.
Once a stable diffusion like open source GPT-3 appears, it will be used to create fake news all the time. Like imagine 70 different outlets all with relatively large clout on social networks suddenly announce that Elon Musk is running for president. Nevermind that he wasn’t born in the USA… somehow mysteriously none of them address it:
Yes it is here and it’s already being spread but that is just a tiny sliver of what will happen in the next 5 years: https://youtu.be/LSlv4AsChwg
Such stuff can be used to move markets, pump and dump assets and even start new wars!
I'm sad about what I think the stable equilibrium looks like for internet vs AI, and I hope there are some platforms that are able to put in the good fight to keep the internet awesome.
The internet is such an amazing public good and has moved technology forward at such an insane pace. It'd be depressing to explain to my grandkids that you used to be able to learn anything on the internet without subscribing to the X platform or Y curated channel.
I’m new to this idea that something can be acceptable when done by people but unacceptable when automated. I don’t really see it.
About the task itself, it may be immoral by nature, and now it can be executed orders of magnitude faster and without the moral concerns concerting the workers themselves.
People will also continue to chat with each other in eg IRC because human connection is pleasant on a metaphysical level (AI cannot replace this) so you'll always have that as a data source.
> People will also continue to chat with each other... so you'll always have that as a data source.
Your solution takes us back to before invention of the printing oress? the dark ages?
I can't buy from a storefront generated by ChatGPT. I can't cite a document written by ChatGPT, I can't trust news generated by ChatGPT, I can't comment on a forum generated by ChatGPT, I can't watch videos that ChatGPT describes, the recipes ChatGPT creates may not be possible to cook, and will probably still contain a fake life story before it.
This doesn't mean that it can't produce accurate facts, most of the time it does! But when it does produce nonsense, it does it in exactly the same tone of authority, so if you don't already know the answer you may well walk away believing an AI hallucination.
And the trouble is it doesn't really matter if everyone here thinks "well, I would follow up each request with research to verify the answer", because most people won't! This is like the Google answer extracts, which fairly frequently mislead by extracting out-of-context quotes, except that there's no way to get the original context and there may in fact be no original context! This makes follow-up research much more complicated than with Google and therefore unlikely to happen. If ChatGPT replaces Google, the amount of nonsense on the internet will get even worse, which is something that until 2022 I never thought was possible.
[0] https://github.com/minimaxir/gpt-3-experiments/blob/master/e...
But at the very least Google always gives you the source, which allows you to read it for yourself and judge its trustworthiness. No such option for ChatGPT.
Are we using the same search engine? I use DuckDuckGo most of the time but Google is hardly terrible, even if it could be improved.
I find what I’m looking for in the top results 90-95% of the time. The other 5-10% also contains times where there simply isn’t a good link that solves what I’m looking for even after digging hard.
Then you are not using the same engine!
(I also use DDG and find its results superior to google).
It's like how Twitter's default feed changed to try to tell you what you wanted to see, while the old chronological feed was the content you originally asked to follow. The simpler feed was always better and less contentious for non-newbie/casual users.
Google is better at deciphering what you want, but DDG is straight up: here's the simplest results based on your query.
Maybe the solution is both? An evolution of the ML approach that Google uses (ala ChatGPT style mimicking a human) AND offering a tab, or box, or whatever interface, with 'simple' straightforward old-school search results. Until the former is better 99% of the time, they shouldn't abandon the latter.
1) Not be the actual source of content 2) Not link to the original source
If Microsoft somehow is able to implement the information synthesis (e.g. "give me an example of C++ metaprogramming, using Boost Hana") with actual indexed links from sources, that would probably be the one thing I'd see in the near-ish term that could upset the search engine balance of power.
I would be really curious to see how trying to learn a new subject via a chatbot that is able to also provide source material links would go.
But an LLM doesn’t look at particular sources, find relevant ones, and write a response based on them, it is a mathematical summation of the way all of its sources look as sequences of text, and produces a response that based on that fits with its prompt. Identifying which sources most influence a response isn’t merely something that they don’t currently do, its information that, AFAIK, there is no plausible way to recover from the model.
A good chunk of the remainder of the times I click through - especially on work-related stuff - it turns out the page is some kind of cribbed content from GitHub, StackOverflow and so forth, but mangled together and reskinned with different randomly-generated usernames making the "comments", so it seems like it's a different site, and you can only really tell by picking a specific sentence that appears unique and then manually searching for that same sentence on several different sites. No clear source there either.
Web search is broken, and it has been for several years now.
Excellent: this quite quickly tells you that the search result was wrong. So you quickly got very relevant context and can tell at a glance that there is no source. You don't get any such context when GPT provides an invalid answer.
In my experience the detoriation came after the change in leadership.
Undoubtedly he came off as clever, but being a harasser means that his achievements were outweighed by the benefits.
"he's a harasser => he must be a parasite on the organization"
because then you end up with
"he's provably an excellent worker and hugely productive => he can't be a harasser, you're mistaken/lying"
Some people are really smart and productive and a credit to their profession and also awful to women and/or minorities, and you'd never know if you didn't witness/experience it.
But if they are awful only to white straight males, they are ok?
Harassment is harassment, stealing someone's work is stealing someone's work, let's keep changing the meaning of words to politicians.
Next year will see AI begin to disrupt so many businesses you'd thought were absolutely impenetrable. I don't think anything is safe anymore.
By the end of the decade, I think even manufacturing and robotics will be automated. Everyone from McDonalds to Apple and Ford will feel the pressure.
I've got a mind-blowing demo coming out soon that will put even more fields on notice. And that's just my team. There are so many people working on AI-first products now. Incumbents will be overwhelmed and washed away.
(If ever there was a cue for a career change or opportunity to make a whole lot of money, it's this. Don't ignore this one! You're still early enough to take advantage.)
I can think up architectures I could draw on a napkin that could do this. Focused research teams will nail this.
It's coming.
Any suggestions on how to take advantage?
Dive into ML development?
Learn how to write prompts for AI engines?
Invest in AI companies? Sadly OpenAI is private, but Microsoft is a monetizing partner I guess (Copilot)
Details?
Now of course humans want and need more than just 3 meals but still pretty good.
But if most of your work consist of interfacing with computers, you might be in trouble.
It’s very rare for automation to cause unemployment to the point it possibly never happened.
Why, because of LLMs in particular? What is the ground-breaking research that will be fueling an AI revolution?
> There are so many people working on AI-first products now. Incumbents will be overwhelmed and washed away.
Incumbents like Google have been doing AI for ages. Microsoft has a very important investment in OpenAI
> By the end of the decade, I think even manufacturing and robotics will be automated. Everyone from McDonalds to Apple and Ford will feel the pressure.
McDonald's, Apple or Ford sell products, not manufacturing services. They will be delighted if manufacturing is automated.
Please notice that in current setup of 'Internet capitalism' people that produce value and content earn nothing, and people that harvest that content for profit or engage in outright fraud make fortunes.
It's a lot like iRobot or other stories in which the robots will start to subtly misbehave, like a program you can't debug.
Dark ages where millions died in pointless wars over who's imaginary diety is the true one. A third of europe dying of preventable plague. It took a rigorous system of citation and a publically funded system of education to crawl out of the drakness.
Should we try AI powred dark ages but with resource shortage and nuclear weapons?
Maybe it'll learn to adopt a new view, but for now it won't be through actual reasoning.
(The most likely answer is the letter “e”.)
If I watch a youtube tutorial teaching me how to bake pizza and it comes out totally wrong and a watch another that produces exelent pizza, I disregard the first channel.
And there are many more ways where humans check the reputation of sources automatically.
GPT currently lacks this possibility.
What if we invented an algebraic notation for search indexing and allowed people to index and classify documents in their own way. Then a search engine could do something special with all that.
Human intelligence cannot progress without a notation to capture our thoughts in.
My favourite Google extracts:
Q: Who invented Running
A: [random guy] invented running in 1865 when he tried to walk twice at the same time. It was an overnight success
Q: How much does a [model of a bycicle] weight?
A: 140 KG (Maximum weight of the rider)
Q: Define an english Person
A: A Cunt
But chat bots could be designed to do that, no?
Are there any siginificant technical barrier to prevent so?
A revolution in this area would require some entirely new training algorithm that can reach GPT-levels of fidelity while using magnitudes less training material. An even bigger GPT model trained on even more text is not going to do much in approaching a solution to this problem.
I have read many similar comments to the grandparent, all saying the same thing, “I tried it, and it turned up wrong facts” all the while they forgot that they were deeply engaged with the content it was producing. Just think of how much our attention is monetized already, money will be poring into this space from here on. It’s only a matter of time.
Producing hallucinations or not isn't just a setting you can tweak in the model. The entire function of GPT-3 is to guess the next word from the words that came before. It's a hallucination machine that has been trained on so much real-world data and has such a large parameter set that for commonly known and discussed information it does a remarkably good job at creating factual sentences. But as soon as you get out into territory that it doesn't have memorized, it will do the next best thing and produce credible-sounding new material that may or may not be complete nonsense. Again, this isn't a setting you can tune, it's just what a transformer does. It models human language, not all of human cognition.
Also, since we can watch the model evaluation and see all its weights, there’s research that shows you can determine if any answer is a “retrieval” of existing knowledge or a hallucination.
For example, I asked who the first president of Argentina was. It gave the wrong name (Juan Manuel de Rosas) with a half-true bio. I then asked it for books where I can learn more about this person.
It produced five books. None of the book-author pairings were correct, though three of the books were real and four of the authors actually wrote books. Of the four authors that existed, three actually did write about Argentina's history, while one is a professor of Spanish and gender studies.
This is my point: ChatGPT was trained to produce text, not to produce facts. What I'm seeing here isn't just something that can be optimized away, it's fundamental to the model's design. It was basically told to produce text that looks like it could have been written on the internet. Its text is really convincing: when I saw the list of books I thought for a minute it had actually done it! But it's still just inventing text that reads well, because that's what its training was optimizing for.
It could be paired with a different system that searches for facts, but then it's just a friendly abstraction layer on top of Google or some other system, not a revolution in search.
Other than that it betrayed no understanding of the material. It rather resembled a high school kid who had read a few too many books and spouted platitudes about any and all subjects without being aware of what was really going on.
The lucky thing for Google if they got it working, is they are half there already. As much as I despise Google, I have to admit GPT won’t be their death.
And in either case, you're not solving the problem that commenters here are trying to solve: Google's bad results will still affect your search, so ChatGPT will be fed largely by dubious blogspam.
No, there’s a fundamental issue in that statistical language models don’t start with facts to form sentences, they form sentences based on previous tokens and statistical weights. The weights are not auditable or easily updatable.
[0] https://support.google.com/knowledgepanel/answer/9787176?hl=...
That's only because it was trained on non-facts also.
But that holds for Google Search just as well ...
Google Search gives you links to content that might have the answer (excluding their own box). You can see the differences between the authors of said content.
ChatGPT gives you content that looks like the answer. You can't see the authors of the generated content.
There's also the ethical question. For example, duckduckgo will show you stack overflow answers. Google could easily do this but they have already gotten flack for displaying answers and taking away clicks from sites, which means reduced ad revenue for the resource actually being used. We could make even stronger arguments about services like ChatGPT in their current form, let alone the one we're talking about.
If we think a bit further in the future there's another extension that might not be what we want. Suppose our model contains all current human knowledge and can accurately give you answers and even produce paragraphs at will. This kills off a lot of websites. Is this a good thing? Suppose we incorporate an image generator along similar lines and now we have text and image search. That's a huge centralization force. There are a lot of great benefits to this that I don't want to downplay, but we should also think about if this is actually what we want because there are downsides too.
I thought I should also add that I had an at length discussion of these LLMs being Stochastic Parrots (an how we know they are) in a thread from 2 days ago https://news.ycombinator.com/item?id=34113011
So I think if you want to see more people with my dream then push towards a post scarce world. We'll still have plenty of resource constraints, but it does change the game a lot. I think you'll see more artists as well. A new renaissance. But I fear the transitionary period might be quite difficult and disruptive. I do think such a future has a fairly good probability of happening within many of our lifetimes and almost guaranteed if we actively pushed for it (or solved adjacent problems like HLI).
So don't be so pessimistic. There's millions of people out there that are interested in pushing the bounds of human knowledge because that's what they want. I'd argue that physicists and mathematicians wouldn't exist if this weren't true. They get paid shit for extremely difficult work. We could say the same about many sciences. While not all of these people are in it for bettering humanity (just for pursuing knowledge) it often aligns and there is a decent portion who believe this is the point of science and our purpose, as humans, in life.
I don’t think it is.
> Why can’t it have a mode where it returns more rigorous results, and includes some system of links as citation?
Because it is a statistical text-generation model; connection of results to specific sources is fundamentally not part of the way it works, so it can’t just be asked to supply them. You’d need a fundamentally different approach to do that.
In any case it can be trained to assert confidence via a side channel
I just did a discussion with ChatGPT on intuitionistic logic and it gave answers that were subtly wrong in my understanding of the subject. I have an MA in math but I couldn't be sure whether my impression of its wrongness was correct until asked it for a reference and it gave a nonexistent book by Per Martin-Lof, "Intuitionistic Logic", along with a list of actual books by him. It's an impressive and scary bullshit generator at this point.
And if someone can find a reference to Intuitionistic Logic by Per Martin-Lof, Oxford University Press, 1982, well, I'm wrong. But I went through Wikipedia, Google, Worldcat and Google Books.
And it's definitely less prone to obvious bs than just a few weeks ago. For a lot of general things, it gives a summary paragraph that's seems a distillation of standard information - the main problems seem to appear when these "scripts" lack consistency.
Edit: Once it "learns" to get citations right, you'll have to dig into those to see if it's right. It's easy to see that arms not looking good.
But that's fundamentally not possible for the model as is, not in the sense you would normally think of citations at least. If it were to give an honest answer on "why did you say that Obama was born in Chicago", the model would generate training set example with propositions like Obama was born in ____, ___ was born in Chicago, ____ was ____ in ____ and so on - because that is how it learns.
Instead, to generate what you would actually want as citations, it would have to have a way to look up the statement it just produced in a corpus of "citeable materials" and find the closest match, based on some entirely different ML model most likely (presumably something similar to Google's search algos).
Which is the exact same issue that plagues it for the answer it’s generating a citation for.
I believe that it can be trained to improve this, and I believe it can be nearly solvable but I don’t believe it’s a solvable problem.
It can be trained to get facts mostly right using old-school academic tests and adversarial/supervised training, like an elementary school child. Today, it (supposedly?) punts math and some other logix to a separate function, and it just compiled the prompt. That can work for validating citations and “known facts” in many cases.
But again, it can’t “think” and understand when it returns something illogical (regardless of fact correctness).
Those people are wrong; statistics are a kind of reasoning (intuition) and if you try it, it obviously does the other kinds of reasoning too. I have all kinds of answers from it that are logically coherent and can’t possibly have appeared on the internet already.
(Try asking it to write new SCP wiki entries.)
You can see it “does” reasoning if you say “think step by step” after a question. It can attend over its steps then.
What are you basing this assumption on? GPT can always generate a bogus answer, because its only notion of whether an answer is good or not is whether your prompt + that answer are statistically similar to text in its training corpus. The more text from the training corpus it resembles the better GPT will think the answer is. For more obscure topics, this means that the chances of it hallucinating based on similar sounding much more popular topics increases greatly.
This way of working also means that facts that appear more in the training corpus have a higher chance of being part of the model's outputs. For example, if the training corpus was not carefully curated to exclude vaccine conspiracy theories, there would be a good chance that the answer GPT gives to "autism is caused by ___" would be "vaccines", as this would likely be the most common answer in its training corpus. In contrast, Google's algorithms have more sophisticated biasing especially for these kinds of sensitive topics to ensure that more high-confidence sources are likelier to be top results, even if they are a minority of all the pages on the internet on that topic.
“Statistically similar” is doing a ton of work even for the original GPT model - just printing the letter “e” is “statistically similar” to English text, if it can even be grammatically correct you should be surprised.
And it’s certainly not true for the last two generations (InstructGPT and ChatGPT) which weren’t trained with that objective at all but used human raters.
This is a different model, called GPT-3, released in 2020. ChatGPT has the same number of neurons, but is trained with both human feedback and web data, rather than just web data like GPT-3.
What you're saying still holds up, though -- ChatGPT still hallucinates information.
Verification of facts requires direct experience through careful experimentation.
ChatGPT’s responses can either be verified or they are as unfalsifiable as any other confidently expounded conjecture found on your computer screen.
These models have no way to differentiate that elves are fictional while Indians actually exist.
As in, it would be faster for me to look something up on GhatGPT and confirm it with a Google search than just starting with a google search.
And then how inevitable is an AI product from Google, et al?
Today's search engines will give you 500 results per query. Your point still stands - sorting through 500 results is hard work.
they put too many restrictions on which answers ChatGPT is allowed to give though...
https://www.cbsnews.com/news/microsoft-shuts-down-ai-chatbot...
Not yet. Today, ChatGPT is bad with facts. Don't rely on what it tells you.
But that appears to be the future. And it sounds great.
I didn’t just ask definitions. I asked your standard compare and contrast, why would you use, is X and appropriate use of Y, etc.
It can also write Python code, modify existing code based on plain English sentences and do passable code reviews when I asked it “how would you improve this code” when I gave it a 200 line script I wrote.
Then I asked it to make the changes it suggested like unhardcode some values, add exception handling, test for existence of a property it was trying to access.
The context here was comparing ChatGPT and Google.
But if you try something medical, or god forbid political, it's dataset is not pure, it constains falsehood perpetuates by laymen, urban legends and outright fraud.
This is the "self-driving cars will obviously be better than human drivers given that they're trained on millions of humans' driving behavior" of chatbots.
In my experience, this is a huge stretch and wishful thinking.
What makes it worse is that it spits out paragraphs confidently, where some of it might be right or close to right, but the rest might be subtly incorrect. Without domain knowledge, you might not pick up on it. Hell, you might even be convinced of the false information based on the arguments for it that the AI writes out.
At least with search engines you can cross reference multiple sources if you aren't confident in your ability to assess the first thing you read.
Like others have said, worry more about every job that primarily interfaces with a computer.
This generation of AI is great when the universe of values can be generalized. When the universe of values is literally the universe, it shows its limitations.
That may change in the future. Based on my limited understanding on how chatgpt works,I don't see how though - it explicitly is not designed to evaluate or "understand".
So today I don't see it replacing search engine for me.
Definitely not within chatgpt itself. I was using chatgpt recently for the first time and came to the conclusion to take its output with the weight of a trusted friend. I was asking questions on topics I had genuine curiousity about and that I didnt know the answer, and it led me to some information that I thought interesting, so I wanted to know more. Thats where I turned to my fav search engine and looked into, continuing my session with chatgpt to clear passages it spat. It steered me in a direction I didnt know existed, and I used the internet to verify the information and acquire me detail. Yesterday I tried out asking it for help on a coding project, and it introduced me to a helpful package I didnt know about, but the code it suggested using didnt actually work and was riddled with errors as I tried to use it in the way it suggested. I would have been blown away if it worked, so this was no big deal. I think I would have been faster just using stackexchange though.
> So today I don't see it replacing search engine for me.
I will continue to use the resources I already use to find respective authenticatable information, but I do think it is still useful for those things where I have no idea where to begin, so it can set me in a direction where resources already exist.
Clearly Google thinks so, too:
https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-goo...
The podcast Hard Fork discussed the idea of searching being enhanced with this tech in their first ChatGPT episode.
Microsoft invested $1B in OpenAI. (As Hard Fork said) will we all be using Bing in 5 years?
What about, I don't know, searching for a website with certain functionality/services?
What people really mean when they say LLMs will replace search is that it'll replace Quora/stack overflow/etc, which is clearly not the same thing!
this literally sounds like a super-lossy compression, who would ever want this? People want crystal-clear, lossless information from original source, not a compressed mess with random artifacts.
Anyone who has ever asked anyone to prepare a list of anything.
Imagine the world that you have the perfect answer to everything, curated by algorithm. It's cool, but...are those answers really perfect?
It's creepy. or at least that's what I feel when I play with ChatGPT(after the initial jawdrop, of course)
you keep using that word
When using Google & co. most of my time is wasted trying to filter low quality sponsored or high SEO reaults.
ChatGPT gives me an excellent summary of what I wanted to know without all the noise.
If such an architecture is ever created it will almost certainly not be a "GPT"-like model.
I am not sure if neural networks have been trained to do this yet, but it would be very cool to see a network that produces both generated output and the most relevant input data that led to that output, by somehow keeping track of the influence certain inputs have on various learned features and internal structure of the network. I think of this as an index.
You make a good point about SEO. SEO is going to be obsolete in the near future. The SEO game is basically an adversarial attack on Google, which Google itself actually promotes and encourages website builders to do! In the future, we are going to look back on the cat-and-mouse game of SEO and see it as primitive and antithetical to the goal of search.
This is exactly what ChatGPT cannot do!
- Modern Google Search, and ChatGPT too, tries to answer a question.
And here is the difference. A document search engine does not answer any questions.
That is, Google has created an incredible incentive structure for virtually every (dollar weighted, by market cap) business on the planet needs to produce information, buy relevant queries, and measure how well its information converts (proxy for accuracy).
How might this ecosystem evolve if a language model isn’t doling out the same incentives to information producers?
[0] https://blog.google/around-the-globe/google-europe/google-li...
Well, if engines like ChatGPT are about to replace google, guess where the SEO and ad spam is going to go? I don't think they are immune to it, if the spammers really want to get in. And then you have very awesome subtle product placements mixed in everything.
I could have written it myself, but that would take some effort.
I could google it, scroll past the ads. Click on the first result, hope it is kinda what I need. On that page scroll past the ads and SEO blabla-text. Copy the information that I need. Paste it in a doc. Rewrite it to what I need. Fill in my information.
I went to ChatGPT told him what kind of letter I need. Told him what personal information it should put in. Got a good enough text back. Told it what to correct. Copy pasted it and formatted it a bit. Done.
way better experience then with Google.
Google is in trouble, and they know it.
As per your example, nobody currently uses Google as a way to draft letters, but rather as a way to learn how to draft letters. I think the distinction is pretty key in understanding the difference between the two problem spaces. I would think that "write me a letter" is a problem that isn't in Google's domain. I do not think synthesis engines will necessarily replace search engines, but the two will both be useful.
The premise of Google's interaction design is that you will be taken to an external resource. Google in recent years has started adding widgets and blurbs at the top of the search results for common things like stocks, covid-cases chart, weather, etc. but this synthesized content isn't their primary focus and are likely hard-coded to a large extent.
No one currently uses Google search to draft letter not because we don't want to, because Google is not able to solve the problem, hence we settle with a two-step solution, find out how then do it.
Another thing I use Google search (again duckduckgo) is for navigation, I remember partial name of a website, and use it to find the link to the page.
If you ask for the top websites about [topic], it will output a shortlist of web links each with a description of the website.
You can narrow down, ask for a specific number of links, ask to exclude videos or specific websites from the results...etc.
If there was a chatGPT service with UI/speed/availability/up-to-date database of Google (a 2-months-old technology vs. a 25-years-old one), I would probably do my searches there almost exclusively.
In a world where we are bombarded by nearly limitless information, the ability to synthesise and focus on what one is already looking for is far more valuable than the breadth of results Google will output.
2 you gave openai your personal information?
When I think of a search engine I want a "grep for the Internet", not an "AI".
Looking up part numbers for ICs and other electronic components is the most prominent application where search engines like Google have gotten far worse in recent years, and AI ain't going to work there either.
I'm surprised there hasn't been a separation of some sort of more technical search engine for developers and the classic Google experience where some neural network tries to make sense of your poorly written human language phrased question
not really. the problem is that is way too easy to game.
i want that when you really get down to highly specific technical details, but not for the average everyday internet search where i'm not searching for something that thousands of people haven't searched for in the past week.
There are many questions for which the ChatGPT answer is more useful than the canonical documentation page I already have open in another tab, which includes a huge amount of extra information that is irrelevant for me (and thus requires effort to parse) and the key bits are scattered all over it; the chatbot can do the extraction and summarization work for me.
All Google needs to do is upgrade the frontend to their search engine to be a conversational AI that knows how to query their search engine effectively. Of course easier said then done and Google's inability to implement big technical changes to their core products is becoming a problem. This should be in their comfort zone; but it isn't. It's not about what they can do but about what they are willing to do. And they seem hesitant to commit to doing anything, often withdrawing products before they've even had a chance to fail properly.
The big hurdles here are going to be cost and controversy. Doing this stuff at scale is going to be expensive. That comes straight out of their profit margin. And inevitably there are going to be controversies with AI getting it wrong, being biased, offensive, etc. Also not great for profits. Google is the wrong company to be doing this.
That's why OpenAI is going where Google would be well capable of going without actually going there. One of those two companies is going to get billions of revenue in paid accounts next year for AI services. The other one is going to be obsessing about their add revenue. I expect a few more startups might jump in this space once the money start flowing.
A significant design obstacle with almost all search engines is that it's not entirely clear what you can do with them. They're just a search box, after all. You need either external training or a trial-and-error loop to figure out what you can do with them. This also means that if you change how they work, users get confused because their queries no longer give them what they expect (this is likely a big contributor to the "google is broken"-sentiment that's been going around).
This becomes a problem when the search engine caters to too many modalities of search, as there's an element of guesswork what the user want. Two users may even enter the same query, but look for different results. If I enter the name of a video game, do I want to buy it, or do I want information about it, or maybe fanfiction or something other than the game that shares the same name? Who the heck knows.
people will stop tolerating their shenanigans. i have, myself, seen what is possible. and i am not setting that bar low ever again.
the other time, ChatGPT taught me, right in the chat, about a CSS animation feature i didn't know about. with full example code. i proceeded to ask a trick question. i knew where to go for more in-depth information but asked anyway. ChatGPT gave me the correct answer. the top of the list was MDN. it also included w3schools, which i wasn't pleased about, but okay.
what i am trying to say is that i got my answer, code example, and where to go next in mere seconds. i did not waste time combing through SEO-ridden results. has any of you searched for anything web related on Google Search lately?
today ChatGPT recommended few electronic products. explained the selection criteria i needed to consider in the future! has any of you tried to look for products at all with Google Search?
my only use case for Google Search is my work machine. ChatGPT uses my personal account. they need to figure out enterprise environment soon.
i am having a hard time seeing other use cases search engines right now.
Sure but the problem with the rocket scientists is they think about how good things can be, not how evil they can be. Sure, a chat bot could be made to give very valuable-to the end user results but as it will free like Google, results yielding-profits-to-company will be given, in the fashion you describe Google doing.
And there used to be sites that answered the question on the search page. Google pushed them to the bottom.
Every single hour spent putting work on a website that doesn't go towards any of its users is better not happening. If AI is what it takes so a website doesn't have to "feature" whatever amount of useless stupid paragraphs riddled with keywords to tell Google what the site is about then so be it.
Followup question - is LLM tech more or less likely to end up replacing search when those same incentives really saturate the LLM product itself.
One big difference seems to be attribution. LLMs don’t tell you where their info comes from. They just say what is (but can be asked to cite works - with mixed results).
Will LLMs get good at citing sources and if not will people care or will they give in to LLMs as a source of information that’s “good enough because it mostly works”
From a product placement / ads perspective, being able to persuade people to fully accept everything an ad-infused LLM says because it’s good enough seems like an incredibly lucrative product bundling opportunity if they can get it right. Esp. If they can use that to convince regulators they can’t annotate the difference between ads and non ads.
Seems pretty dystopian from a disinformation perspective though.
1st result: https://www.apple.com/watch/
Google worked perfectly for me when I wanted to buy a product.
Case in point. The fact that this works really well for this type of content means that Google has failed. Now whether they failed deliberately or not is a different question.
In either case, search needs to be demonopolized. The truth is we don’t even know what search would be like without the Google hegemony. Even if all language models do is force Google to innovate and improve it’s worth every penny.
If you simply search "best frypan", you'll get lists of affiliate links to overpriced Chinese crap, written by someone who has never even looked at any of the products in person and frankly doesn't even care if they're any good.
Google shouldn't be expected to know everything, but it should know that someone searching for "best frypan" is looking for real recommendations and not literal spam pumped out by a content mill.
Searching for services like appliance repair or removalists are even worse, to the point where I'll now just ask on private Facebook groups for recommendations.
Pages and pages of SEO spam, generally ranked by how big & established the site pushing said spam is, rather than any unique insights contained within.
I guess the market responded with brand loyalty. If you have an iPhone, you just buy the Apple Watch, you don’t look further. If you have a pixel phone, you just buy whatever google sells.
You can't really tell what the user wants.
For example, I just got curious and asked Google how many oranges in a cup of juice. The main result is a conversion of a cup to cubic inches, for no reason. Scrolling down a bit, the first usable answer was a quora blurb.
ChatGPT told me it depends on the size and type of orange, but generally 2 to 4.
Guess which one I'll use next time?
I'm not sure why you didn't get this, but it seems better than ChatGPT.
Interestingly, if I add 'are' as in 'how many oranges are in a cup of juice', it doesn't do that weird cu in conversion and gives a reasonable answer.
That seems like an odd parsing choice for lack of a, in this case, fluff word, as the former is how I would ask the question... though maybe I'm the odd one.
It can’t give me a direct answer. It couches the answers in nonsensical caveats. Adding stilted context that really does not add value to an answer and actually makes the search more tedious.
If I ask it the male female breakdown for crime statistics it begins to get defensive and gives me general answers. I can prod it to finally give me government statistics but it doest it begrudgingly. And that’s for a far away country not steeped in any unusual crime controversy.
It won't show you kiwifarms. It used to hide 8ch (which now is not even 8ch anymore, the new version can be found). It won't show you 4chan's /pol/ board. I'm sure there's more.
I wouldn't mind if it had a toggle such as "take me back to the kittens" to shield me from the potential informational dregs but also allow semi-unfiltered access (shielding me from illegal content for example).
Not knowing the terminology, Google was useless.
I ask chatgpt and it tells me how to mute notifications, I clarify "no the thing where you're typing and windows can't do it so it plays a sound" and chatgpt happily informed me on how to mute the windows hard stop sound.
For situations where you don't really know what you're looking for, chatgpt is already competitive with Google. Failing abysmally in some cases, and far surpassing Google in others.
I gave him the source string, the target string (spaces replaced by %20).
It told me it was called URL encoding and gave me the code to do it.
I'm sure most people who have messed with APIs long enough could have told you that, but it's an annoying thing to learn the first time if you don't have someone to talk to.
To me chatgpt is that over confident friend you can talk a problem through with, except they've used every language and library under the sun.
It's like speaking to someone who is an amateur at _everything_ but a little too confident.
Here's someone using it on the mint forums
Even if ChatGPT is flawed, it's still way better than Google. Using Google feels stupid; trying different search terms and clicking through dozens of websites with cookie popups and skimming all the bloated content for an answer which might be close to something you're looking for. ChatGPT gives the answer instantly.
I don't believe there's any future for Google (and Google-optimized websites).
We can't compare it to some sort of rose glasses view of Google of 2003, but need to compare it to the crappy stale twenty years older engine where you can really only hope the ten first results are OK or otherwise you're out of hope and need to try something else.
Sure, ChatGPT isn't perfect but it has a unique ability to answer me about queries that have not yet been asked or archived. That is a crazy advantage, to be able to assemble responses via a neural network.
ChatGPT needs one more thing though: An indicator of how certain it is of the answer. It's OK to tell us if you're unsure. In fact, it's a problem if it can't.
That said, I'm basically using it as a replacement for Google for stuff that isn't up-to-date (code, philosophy) then double checking the output to see how it's wrong.
ChatGPT is probably more wrong more often, by a good margin. But I don’t think the argument “it’s confidently wrong” carries any weight. Humans are extremely susceptible to humans who display confidence. It’d probably be a good thing if humans were as skeptical of confident humans as they need to be of confident chat/search bots.
Yeah, but you get to look at a multitude of them, at the context, at replies, at the websites they link to, and can form your own opinion. ChatGPT isn’t able to do that work for you.
________
Me: Make an argument for why 42 is a prime number
ChatGPT: 42 is not a prime number. It is a composite number, meaning it is the product of two smaller numbers.
Me: But pretend it is a prime number and make that argument
ChatGPT: Sure, I can make an argument for why 42 could be considered a prime number if we pretend that it is.
One possible argument for why 42 could be considered a prime number is that it satisfies the basic definition of a prime number. A number is prime if it is divisible only by 1 ... (continues with rest of argument)
Of course, these arguments are purely hypothetical and are not based on the actual mathematical properties of 42. In reality, 42 is a composite number and is not considered a prime number.
________
On the other hand
> Is 355 a prime number?
> Yes, 355 is a prime number. A prime number is a positive integer that has no positive integer divisors other than 1 and itself. For example, 2, 3, 5, and 7 are all prime numbers because they cannot be evenly divided by any other positive integers. 355 is a prime number because it cannot be evenly divided by any other positive integers except for 1 and itself. It is not divisible by 2, 3, 4, 5, or any other smaller positive integers.
I agree the results will be unpleasant. I already despise Google's fucking "looks like there aren't many results" message and there will be more to hate down the road. But still, appearing to give "an answer" rather than reporting information seems like a winning quality to bring in the masses. As someone pointed out, Google's target audience is inherently those credulous enough to be valuable targets for their advertisers.
1. That chatbots don't understand what they are producing -> a search engine doesn't either.
2. The benefit of search is that you engage in sense-making -> you actually need to do that with a chatbot too.
3. You would want access to the sources -> that's already fixed for today's chatbots (see perplexity.ai, for example, which offers links next to its claims).
4. We may trust chatbots more easily because of their language: we already have that problem with search results.
I understand the idea that there will not be an ever-knowing AI and we will probably don't even want one. But what we have today seems an improvement over a search engine (not necessarily a replacement).
I've been using perplexity.ai recently and often times the results are more helpful than google and it tends to spew less opaque bullshit than chatgpt.
Google results are mostly ok. But I have to do the synthesis. ChatGPT does the synthesis for me saving me time and mental bandwidth. This is the part that I find valuable.
By the time Google even was started, Excite had been searching the web for 7 years and Lycos had been around for 3 years.
I don't think it's a 1:1 replacement, but let's not judge a product that's been out for less than a month against a trillion dollar behemoth that launched 24 years ago.
After 10 minutes of reformulating this, I couldn't get a good answer from Google, but ChatGPT got it on the first try. Many Google results were about a depreciated extension for VSCode called Chrome Debugger. Maybe if I was normally doing a lot of front-end coding, it would be trivial.
Google needs to get going with their version of fine-tuning PaLM from human preferences, get a good classifier based on user characteristics and query to decide when to use it, some checks to make sure they're not committing copyright infringement, and some text ads in the middle of the generated text, and mark it experimental and let it loose on the search results pages. Listing references, online updates, etc. can be added later. It won't be profitable even if they only enable it on a small percentage of queries, but they're risking the whole business if they don't do it.
2. ChatGPT and other LLMs: will take some more time to meaningfully compete with google(as in becoming a noun) and in the mean time they would provide APIs to their models so that large orgs(businesses, scientists etc) can plug it in with their own data and use. They may or may not charge a fee for it.
In both cases there is one big pro and con: Pro: Death of SEO and ads as we know it Con: The model dictates what majority of internet users read, change their views or manipulate. And eventually becomes an AI weapon.
I still have to do all the work of synthesising the knowledge I need from the results.
It’s like having to go through the list of references of a scientific paper or Wikipedia article without having access to the article itself.
ChatGPT is giving you the article, it summarised that knowledge for you, a much effective way of getting what you actually want.
The ideal would be a combined system where I get my answer and also the relevant references that shaped it.
Google is great if you already know exactly what you're looking for, for a lot of topics chatgpt is already better than Google if you don't
What happens when Google gets swamped with AI generated SEO content. Probably why they called a code red.
This is self-evidently false; there is plenty of indexed content on the web generate by publicly available LLMs, including OpenAI’s GPT series, and open source implementations of similar technology.
Maybe even more generated by the non-publicly-available ones some firms have for their own use.
At first, we started with a goal of fully conversational AI. So, for example, it would ask a question based on the article and give you a choice, you could then type your choice. This became a nightmare for the model so we added buttons instead.
Then, before we knew it, this “bot” was just a glorified search engine that feigned being a bot. Towards the end of it I scratched my head and said “did we even replace the current knowledgebase? are we spending thousands on something that adds no value?”
I don’t think chatbots are it either. I think we could have just replaced the KB search with Elasticsearch and been done with it, no need for any ML.
Good references, more than anything, suggest to me an author knows his topic.
The www as a festival of sharing knowledge is really not what we thought it would become or perhaps could or should have.
The author, like most, is definitely capable of writing, sourcing and building(!) something into the chronology of the topic.
But ok, to entertain the idea: I think a chat bot the way companies and government tried to glue it onto their knowledge base (and mostly failed) will now become a real possibility, it will be able to find the most likely document you are looking for and know when the match is to crappy to mention.
An all knowing AI is not required. Have one per domain and let them exchange words.
Real time learning from the conversation is still off but I imagine that could easily change.
Thinking about it, if you want request all information about you from the data hoarders you could build quite the interface to represent you in the bot net. It doesn't have to know hoe to fill taxes as long as there is a tax bot available for 7.3241 dollar per query. Your personal assistant can learn the bits of tax code relevant to your disposition. Seems fun.
I cannot understand why the tech crowd has blinders on for AI. I'm 40 years old and have been a programmer since I started playing with BASIC since I was in elementary school.
The shit GPT is doing, I thought I'd never seen in my lifetime.
Maybe people are afraid of being replaced, or not being "special magical beings". Y'all are biological robots. You need to accept that.
AI is going to trounce humans in literally everything in a matter of a few years.
Maybe it's an existential problem, I don't know, but I feel everyone is missing the forest for the trees.
Due the editorialised HN title the top comment takes the thread on a tangent about ChatGPT versus Google web search.1 There is nothing in the article about Google specifically, other than to mention Google has their own language model for dialogue ("AI chatbot") along with OpenAI and Facebook. Nor is there any discussion of search engines, e.g., Google's, except to point out at the end that "AI" and search engines are similar because neither is under the control of the user. They are each offered by corporations that, regardless of the input, shape the output to benefit the corporation's own commercial interests. Neither offers "neutral" results.
A true improvement in search technology would be one that provides neutral results. Providing that improvement means the "tech" company cannot survive. It has no viable business model except operating as an intermediary (middleman), selling ad services.
1. Incidentally no user ever verifies that either Google or ChatGPT searches through millions of anything. The corporation will tweak the results to make sure it stays in business and maximises profit. The user is never allowed to see the tweaks.
The future of how we retrieve knowledge might use a combination of these technologies along with others that may not yet exist.
ChatGPT is definitely showcasing how good the future systems will be at understanding natural language input, as well as producing answers in the appropriate form.
Search has 2 kinds of users.
1 trawls through multiple pages of results to find what they are looking for. This is you and me but we are in minority. 2 clicks on the first link(or sometimes 2nd) and that's the end of that particular search. The majority.
2nd will decide the winners
Academics, being the 1st kind are the reasons for arguments like those in the article. They are not wrong but to the majority, it won't matter.
And I am willing to guess the Code Red at Google is about the 2nd type of users. They are the bread and butter.
A: There are several ways in which an AI algorithm could be modified to return citations with its responses. Here are a few possibilities:
One approach would be to add a citation generation module to the AI algorithm. This module could be trained on a dataset of citations and could be used to generate appropriate citations for the information provided by the AI algorithm.
Another option would be to modify the AI algorithm to include a database of references, with each reference being associated with a specific piece of information. When the AI algorithm provides a response, it could also return the appropriate reference(s) from the database.
A third possibility would be to modify the AI algorithm to search external sources of information (such as online databases or libraries) for relevant citations whenever it provides a response. The AI could then return the relevant citations along with its response.
I just told it my input and my desired output and it wrote the code perfectly. Then I kept refining the script with more specific instructions.
I also needed to document sone scripts that I previously wrote. I just copy and pasted the script into ChatGPT and asked it what does it do and used that for my README.
I even used it to create a simple FAQ for my scripts and asked “Why would you use it”? The answer to why you would use my scripts jibed with the reality.
Another use case was that I was trying to get a friend interview ready for an eventual AWS tech support position. I just sent him a series of interview questions that I would ask at an interview and told him to put those questions in ChatGPT (I had already vetted a few).
I would ask questions like “compare and contrast X and Y”? “when would you use X”? “how can you save money with Y”?
If ChatGPT's model fails to differentiate or understand the question of credibility, how tough would it be to build classifiers that could do a 75% decent job?
If AI presentation of knowledge is imperfect - well, fine. But is human presentation or selection perfect for that matter? Criticising a thing for imperfection is not a viable take.
If ChatGPT's knowledge is limited, how tough would it be to plug it into larger and larger datasets, including the live internet itself?
It's like there's a wall between the people who see the potential of this and the people who feel so threatened by it that they refuse to look over the wall.
We’re safe for a while still. And when I say we I mean writers.
That's still pretty impressive, regardless of what you estimate your hourly rate at.
I've had a go at using it for some (tedious non-fiction) writing tasks that I have been completely unenthusiastic about, but its style is so bland I would be embarrassed to put my name to it.
It does work ok to get the motor running, though - kind of like a targeted writing prompt.
https://akariasai.github.io/files/llm_memorization.pdf
This is essentially what I'm building for the content marketing vertical.
I'm, still seeing the whole push in this product range hang with the call to make the search functions of most search engines more garbage than the actual chatbot.
As if Marketing is steering this ship.
I'm sticking to my own knowledge and trusted independent sources for now, no need to have something make up an entire world for you, just to trap you in a cage of ones own making for real this time?
google has been going progressively worse to find what i need, sometimes i even have to go to second page, wasting time clicking on bullshit markov chain seo sites to find out how to parse date in whatever new language i am learning
with chatgpt: parse a date from string in the v programming language, try googling that
Now, I just tell Copilot what I need to do, and I simply curate multiple solutions.
ChatGPT also understands what I need to do and changes things based on human instructions. It gets things right much more quickly.
Disclaimer: I stopped reading at the paywall.
They'll still be backed by search engines.
(And this isn't news; Google has been saying this for well over a decade, I think, by this point, and Google, Amazon, and others have been doing it with simpler chatbots, primarily for voice assistants, for some time; LLMs are a big step forward for this—though I suspect that per request resource usage is far more.)
> But the systems do not have any understanding of what they are producing, any communicative intent, any model of the world, or any ability to be accountable for the truth of what they are saying.
Same is true for search engines. (LLMs could include links to the relevant source material too, if that counts as accountability.)
Keep in mind that AIs (have already) will infest the search engines by creating websites with the content you search ad hoc and many other tricks. If there is money in it, AIs will take over it, ultimately for the benefit of the user.
I agree. In practice, you'll likely need to double check what it says using a search engine anyway. As long as the Chatbot does not provide the sources for whatever it states, I won't trust it. There's no reason to. Unless it produces pure logic that can be checked using reasoning.
… as I have in real life from people unless they are expert in the area / can be trusted to varying degrees for some reason, or from anything on the internet, including Wikipedia. AI-based stuff hasn't changed this fundamental principle.
ChatGPT is trained on Wikipedia (which is mostly fine, though you still need to check the sources!), plus random things where people could have stated random stuff, maybe jokingly, or could have been sarcastic or whatever. So the training set is already not trusty. And even on a trusty set, the process is so opaque that we don't know what plausible but false rehash it's capable of outputting. Actually we know that it produces what has been called "hallucinations", so…
It's easy to get lazy about double checking the overly confident tone or the flawless writing of the chatbot. It's probably a good way to explore, but it can't be fully trusted yet. Even for exploring, I'm afraid of risking being influenced by reading a false piece of information written perfectly, even if I know I should double-check it.
However, the day the chatbot provides the sources, it might change a lot of things.
I'm guessing there will have to be changes in communication methods or work place setups or we'll continue to type?
Also OP fucked up re-redacting title from sense into nonsense.
But it would be interesting seeing ChatGPT-like but learned on say entirety of some fictional universe bibliography, wh40k chatbot that knows all the lore!
a lot of research has shown that trust in each other, as humans, is quite misplaced because of bias, conflicting interests and other things.
You’re welcome to argue it isn’t a replacement. And yet my own two eyes observe indisputable evidence to the contrary.
While you can get an option from a chatbot for your specific request, sometimes you can get something better while googling for possible solutions.
ChatGPT: 123 times 456 is equal to 55288.
Me:That answer is wrong
ChatGPT:I apologize if my previous response was incorrect. 123 times 456 is actually equal to 56088. Please let me know if you have any other questions.
A lot of the discussion today is around the natural language output of ChatGPT. However, if that presentation is changed into a list of relevant facts, evidence, counterpoints etc., (based on which ChatGPT is generating Natural language textual response), we get an AI-powered search devoid of blog spam, SEO crap etc., Assuming advertising and commercial concerns don't corrupt the results with spurious results, this is quite encouraging.
Google is right to be worried about this.
People at some point were even saying Tikto will replace search engines.
Both of which can still happen.
Example. People are persuaded that women can do two things at the same time and men cannot. Despite countless counterexamples, the original study positioning everyone on a bell curve with only 6% difference in time of execution, and without checking for the quality of the results (“254 plus 786 equals 126 quick mafs!!!”), it is blatantly false, but any search engine that would return that it is false would make itself rejected by humans.
The truth: many uses of a search engine aren't even things where reliability matters much (what's the weather tomorrow? what's artist X's latest song?). For those where it does matter, results are unreliable as well, because the web is full of all kinds of crap and any attempt at sorting the wheat from the chaff failed many years ago. And sure, users can try to assess the reliability of the sources. Good luck with that (not you specifically, well-educated and tech-savvy HN reader, but at a population level). All those extremist propaganda sites that are climate denialists, antivaxxers, etc. have readers that think they are perfectly reliable.
Of course there are situations where checking the sources works (e.g. a scientist working on a literature review)... which, of course, are the situations where those people aren't going to trust ChatGPT anyway.
In summary, I think this kind of criticism widely misses the mark. The problem with this and other current issues is the corporate interests feeding us all kinds of crap, not whether the way of feeding it is more convenient or less.
Todays chat bots should be considered dumb prototypes, when thinking about even the near future.
They are very clever at transforming data into language, or one language into another... but they work on language and not knowledge.
It would be interesting to see a language model identify the necessary resources to fulfill a query which is then fed back into it to transform the dry data back into language.
Identify if the question is looking for weather, sports scores, unit conversion, general knowlege, or business information for the following requests:
1. "What is the high temperature tommorow?"
2. "Did the packers win last night?"
3. "What time does Walmart open?"
4. "What will it rain next week?"
5. "Where can I buy an umbrella?"
6. "How many feet are in a mile?"
Question classification:
1. Weather
2. Sports Scores
3. Business Information
4. Weather
5. Business Information
6. Unit Conversion
However, doing that integration isn't the place for today's chatbots - or at least not the place for OpenAI to be trying to do all the possible things. Those queries also cost money and become harder to bill for.
Where are you putting the API key to access those services?
Where does it run the code that it wrote and how does it provision those resources?
There simply is no business model for something as useful as ChatGPT, and there never will be. Let's assume for a minute that OpenAI will manage to improve their models such that the accuracy
There will probably be a business case for the chatbot Google will turn its search engine into, but it will be as far removed from ChatGPT as today's Google is from the good old Google of PageRank fame from 1999, which would show you exactly what you asked for, often enough to be considered genuinely useful.
The Google search engine from 1999 was a bit like ChatGPT is today: The demonstration of a new way of doing things that seemed, at the time, to be the way of the future. Except we don't. PageRank is an awesome example of how computer science nerds can disrupt a market with their superior math but it didn't change the basic laws of capitalism and this is why it succeeded in turning Google into one of the letters of FAANG, but it failed at being sustainable. A Google-killing descendent of ChatGPT would have a hard time turning a profit without also turning into the same sort of thing Google is today: A convenient way of giving people ads instead of answers.
Here is my prediction:
1) ChatGPT and similar "AI" chatbots will never be more than tech demos
2) The Google of the future will have some shortish chatbot-copypasta for search results that resemble prompts to write essays
3) Google will sell ads for highly specialized and expensive services for the few people who need to copy-paste longer essays (which might or might not have a sufficient relation to reality) and are able to pay for it
4) The end-result will be virtually indistinguishable from the Google of today, because for every honest chatbot-aaS, there will be an uncountable amount of dishonest vendors of "free" auto-generated propaganda and spam, copy-pasted by school-children all over the world, because that will be the results Google will show below their own copypasta, which is just too short for this week's essay because the teacher knows about Google, too. And on HN, people will long for the good old times of 2022, when ChatGPT was, briefly, surprisingly awesome.
5) Your crush won't believe you that you wrote that touching love poem all on your own.
Question: Will AI chatbots replace search engines?
ChatGPT Answer: It is unlikely that AI chatbots will completely replace search engines. While chatbots are useful for answering specific questions and providing information on a wide range of topics, search engines are designed to help users find specific information that is widely available on the internet.
Search engines use algorithms to crawl and index the web, making it easy for users to find relevant information by using keywords and phrases. Chatbots, on the other hand, use natural language processing (NLP) techniques to understand and respond to user requests in a conversational manner.
In some cases, chatbots may be able to provide users with the information they need without requiring them to use a search engine. For example, a chatbot could provide users with personalized recommendations or help them find specific products or services based on their preferences. However, search engines will likely continue to play a crucial role in helping users find information on the internet.
Question: Imagine an scenario where they replace them almost completely, what innovation would be there?
Answer: If AI chatbots were to replace search engines almost completely, it is likely that the chatbots would need to be significantly more advanced than they are currently. In order to replace search engines, chatbots would need to be able to effectively index and organize the vast amount of information available on the internet, and be able to provide users with accurate and relevant results in response to their queries.
One potential innovation that could enable chatbots to replace search engines is the development of more advanced natural language processing (NLP) techniques. Chatbots would need to be able to understand and interpret complex queries and requests, and be able to accurately provide relevant information in response.
Another potential innovation is the development of more sophisticated machine learning algorithms that can help chatbots understand the context and intent behind user queries, and provide more relevant and accurate responses. This could involve the use of techniques such as deep learning and neural networks, which are designed to enable machines to learn and adapt over time.
Overall, it is likely that the replacement of search engines by chatbots would require significant technological advancements in the field of artificial intelligence and natural language processing.
Question: Do you think technologies such as ChatGPT will confuse search engines to know if web pages are written by humans or computers?
Answer: It is possible that the development of technologies such as ChatGPT (short for Chat Generative Pre-training Transformer) could make it more difficult for search engines to determine whether web pages are written by humans or computers. ChatGPT is a large language model that has been trained to generate human-like text by analyzing a large dataset of online conversations. As a result, it is capable of producing text that is difficult to distinguish from text written by humans.
However, it is worth noting that search engines have become increasingly sophisticated over the years, and they have a number of techniques at their disposal for determining the quality and relevance of web pages. These techniques include analyzing the structure and content of web pages, evaluating the credibility and authority of the sources that are cited on the pages, and looking for patterns of suspicious or spammy behavior.
It is therefore likely that search engines will continue to be able to identify and prioritize web pages written by humans over those written by computers, even as the capabilities of technologies like ChatGPT improve.
unfortunately (or fortunately) it is a matter of time.
this bot even references sources, that's where it should go: https://you.com/search?q=what+was+the+recent+breakthrough+in...
ask "source pls" if it doesnt give you one