Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google scrambles to manually remove weird AI answers in search

theverge.com

Post reply on HN
Print view View original

171–180 of 387 posts

Re: Google scrambles to manually remove weird AI answers in search

#171

Post by verisimi »

jerf wrote: post #27

"Achieving the initial 80 percent is relatively straightforward since it involves approximating a large amount of human data, Marcus said, but the final 20 percent is extremely challenging. In fact, Marcus thinks that last 20 percent might be the hardest thing of all." 100% completely accurate is super-AI-complete. No human can meet that goal either. No, not even you, dear person reading this. You are wrong about som…

100% correct, 80% correct lol.

The thing is that truth/reality is not a thing that is resolvable. Not even the scientific method has this sort of expectation!

You can imagine getting close to those percentages, with regards to consensus opinion. That's just a question of educating people to respond appropriately.

Re: Google scrambles to manually remove weird AI answers in search

#172

Post by wizzwizz4 »

madeofpalk wrote: post #152
lukev wrote: post #115

Earlier quoted context omitted.

They don’t just seem it. They are by design. We talk about models “hallucinating” but that’s us bringing an external value judgement after the fact. The actual process of token generation works precisely the same. It’d be more accurate to say that models always hallucinate.

Yes - this is what i've been saying all the time. The term 'hallucinations' is misleading because the whole point of LLMs is that they recombine all their inputs into something 'new'. They only ever hallucinate outputs - that's their whole point!

Into something probable. The models that underlie these chatbots are usually overfitted, so while they usually don't repeat their training data verbatim, they can.

Re: Google scrambles to manually remove weird AI answers in search

#173

Post by wizzwizz4 »

ta8645 wrote: post #169
bluefirebrand wrote: post #167

Earlier quoted context omitted.

> it will give you a logical stepwise progression from your question to its answer. No, it will generate a new hallucination that might be a logical stepwise progression from the question you asked to the answer you gave, but it is not due to any actual internal reasoning being done by the LLM

So what? You have no way to know for sure if the human you ask the same question, does either. The question that started this thread was related to verifiability. And i still think it is a spurious complaint, given that we have exactly the same limitations when dealing with any human agent.

The problem of other minds is no reason to throw everything out the window. Humans are capable of being conscious of their reasoning processes; token-at-a-time predictive text models wired up as chatbots aren't capable of it. Your choice is between a possibly-mistaken, possibly-lying human, and a 100%-definitely incapable computer program.

You don't know either "for sure", but you don't know that the external world exists "for sure" either. It's an insight-free observation, and shouldn't be the focus of anyone's decision-making.

Re: Google scrambles to manually remove weird AI answers in search

#174

Post by EasyMark »

rurp wrote: post #150
PreInternet01 wrote: post #63

It's debatable whether Google has truly lost the plot because of the "AI wars", but the moment the statement "Bing returns more sensible results than you" becomes verifiably true, it's... cause for concern? The approach that Google appears to have taken, which is to assume that the top-ranked part of its current search index is a sensible knowledge base, may have been true some years ago, but definitely isn't now: fo…

Not only is the current internet 80% spam, it's rapidly approaching 99% thanks in large part to LLMs. At this point I would be shocked if Google had a solid plan for how to handle this going forward as the problem space gets more difficult.

that's the part that scares me. I railed on someone's comment the other day about "indexes will come back into fashion" but the more I think about how much garbage has increased in just the past 2 to 3 years, I think I was wrong. Indexes and forums may be the only way to have a sane net where you can find things. Perhaps communities linking together in a ring like format, a "web ring" of sorts.

Re: Google scrambles to manually remove weird AI answers in search

#175

Post by mvkel »

1. Google announces something that has AI bolted on

2. A VP pontificates about how much work they did to "get it right"

3. An easy-to-anticipate first-order issue surfaces

4. Sundar issues a statement like "this is completely unacceptable. We will be making structural changes to ensure this never happens again."[0]

5. GOTO 1

[0] https://m.economictimes.com/tech/technology/sundar-pichai-ca...

Re: Google scrambles to manually remove weird AI answers in search

#176

Post by ttGpN5Nde3pK »

My whole qualm with this AI integration into search engines: it's a search engine, not a question engine. I go to google to search the internet for something, not ask it a question. IMO, asking AI for something is a different task than searching the internet.

It's sorta the same problem as if I go into a store and ask an employee where something is, and they reply with "well what are you trying to do?"

Re: Google scrambles to manually remove weird AI answers in search

#177

Post by ysavir »

EasyMark wrote: post #174
rurp wrote: post #150

Earlier quoted context omitted.

Not only is the current internet 80% spam, it's rapidly approaching 99% thanks in large part to LLMs. At this point I would be shocked if Google had a solid plan for how to handle this going forward as the problem space gets more difficult.

that's the part that scares me. I railed on someone's comment the other day about "indexes will come back into fashion" but the more I think about how much garbage has increased in just the past 2 to 3 years, I think I was wrong. Indexes and forums may be the only way to have a sane net where you can find things. Perhaps communities linking together in a ring like format, a "web ring" of sorts.

What I've been wanting to see for a while now is a social-network based search engine:

* No pages are indexed automatically. The only indexed pages are pages that users say are worth indexing. Probably have a browser add-on for a one button click that people can use. * You can friend/follow others * Your search results are a combination of your own indexed pages and the pages indexed by people in your network.

Re: Google scrambles to manually remove weird AI answers in search

#178

Post by pankajkumar229 »

Szpadel wrote: post #94
jerf wrote: post #27

"Achieving the initial 80 percent is relatively straightforward since it involves approximating a large amount of human data, Marcus said, but the final 20 percent is extremely challenging. In fact, Marcus thinks that last 20 percent might be the hardest thing of all." 100% completely accurate is super-AI-complete. No human can meet that goal either. No, not even you, dear person reading this. You are wrong about som…

I think the biggest difference with human (and the most important one) is that human can tell you "I have no idea, this isn't my field" or "I'm just guessing here" but LLMs will confidently say to super stupid statement. AI doesn't know what it knows. if you only score where human provide answer, then human score would be probably in high 90s

I find irony here.

Re: Google scrambles to manually remove weird AI answers in search

#179

Post by mvdtnz »

jerf wrote: post #27

"Achieving the initial 80 percent is relatively straightforward since it involves approximating a large amount of human data, Marcus said, but the final 20 percent is extremely challenging. In fact, Marcus thinks that last 20 percent might be the hardest thing of all." 100% completely accurate is super-AI-complete. No human can meet that goal either. No, not even you, dear person reading this. You are wrong about som…

> No, not even you, dear person reading this. You are wrong about some basic things too. It'll vary from person to person what those are, but it is guaranteed there's something.

The difference is that I'm not put on the interface of a product facing hundreds of millions of users every day to feed those users incorrect information.

Re: Google scrambles to manually remove weird AI answers in search

#180

Post by tedunangst »

kibwen wrote: post #105
odyssey7 wrote: post #91

This is analogous to the Apple Maps launch failure. Except that Apple competes to make the best smartphone, and an iPhone was still valuable without Apple Maps. What happens to Google if it stops being able to compete in search?

"Search" isn't Google's product. Google hasn't been a search company for 20 years. "Ads" is Google's product. And the only way they'll go bankrupt is if 1) companies realize that advertising is pointless (I'm not holding my breath), or 2) some other company takes over from Google, which seems unlikely without government intervention (I'm not holding my breath). Google is a shit company, but they'll still be around 20…

Still need visitors to see the ads.
Post reply on HN