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

51–60 of 387 posts

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

#51

Post by b112 »

JCM9 wrote: post #17

The fall of Google’s reputation on ML is nothing short of spectacular. They went from having a near untouchable reputation as being far ahead of any other large tech company on ML to total shambles in a year. Everything they’ve released has been a complete popcorn worthy dumpster fire from faked demos, to racist models that try and pretend white people don’t exist, to this latest nonsense telling me put glue on my pi…

At least Elmer's white glue is edible, millions of kids agree.

(The logic sort of makes sense. Glue sticks things together, and some glue is edible.)

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

#52

Post by krapp »

SoftTalker wrote: post #43
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…

> putting glue on pizza to hold the cheese on It's actually not the dumbest idea I've heard from a real person. So no surprise it might be suggested by an AI that was trained on data from real people.

It wasn't an idea, though. It was a joke someone made on Reddit. If an AI can't tell the difference, it shouldn't be responsible for posting answers as authoritative.

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

#53

Post by moomoo11 »

rvnx wrote: post #26
rchaud wrote: post #21

Earlier quoted context omitted.

...for our shareholders

I hope AI will bring back the "Sort by date" button on Google Reviews, and add somewhere a Google Maps link. Who knows, maybe AI can bring back exact keyword matches, or correct basic math calculations on Google Search too.

It will cost $2 billion of nvidia chips and it won't work.

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

#54

Post by sgift »

SoftTalker wrote: post #47
thesimp wrote: post #37

I'm actually shocked that a company that has spent 25 years on finetuning search results for any random question people ask in the searchbox does not have a good, clean, dataset to train an LLM on. Maybe this is the time to get out the old Encyclopedia Britannica CD and use that for training input.

I am also surprised that training data are not much more curated. Encyclopedias, textbooks, reputable journals, newspapers and magazines make sense. But to throw in social media? Reddit? Seems insane.

Even some results from "The Onion" seem to be in it. Looks like Google just took every website they've ever crawled as source.

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

#55

Post by Barrin92 »

dazc wrote: post #2

Manually removing rogue AI results is kind of ironic isn't it?

generative AI is essentially three day labourers from an emerging economy in a trenchcoat. From data labelling, to "human reinforcement", to manually cleaning up nonsensical AI results.

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

#56

Post by ADeerAppeared »

cjk2 wrote: post #38
ADeerAppeared wrote: post #31

Earlier quoted context omitted.

> So Google hasn't used an LLM to generate and test weird queries ? You don't even need an LLM for that. Google will almost certainly have tested. The test result is just politically-unacceptable within the company: It doesn't work, it's a architectural issue inherent to the technology, we can't fix it. Instead, they just rush to patch any specific, individual errors that show up, and claim that these errors are "rar…

They already know it’s a shit show. They are trying to push it along until it’s someone else’s fault.

I'm not convinced the executive layer is aware how dire the problem is.

On one hand, their support for outsourcing programmes; "Training Indians on how to use AI", suggests they realize AI tooling without human cleanup is a crapshoot.

On the other hand, they keep digging. This kind of gaslighting is an old and proven trick for genuinely rare problems, but it doesn't work if your issues are fairly common, as they'll get replicated before you can get a fix out.

Similarly, they're gambling with immense legal risks and sacrificing core products for it. They're betting the farm on AI, it may kill the company.

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

#57

Post by calebkaiser »

JCM9 wrote: post #17

The fall of Google’s reputation on ML is nothing short of spectacular. They went from having a near untouchable reputation as being far ahead of any other large tech company on ML to total shambles in a year. Everything they’ve released has been a complete popcorn worthy dumpster fire from faked demos, to racist models that try and pretend white people don’t exist, to this latest nonsense telling me put glue on my pi…

There was an interesting interview with David Luan about this recently. For context, he was a co-lead at Google Brain, early hire at OpenAI, and is now a founder at Adept: https://www.latent.space/p/adept

The TL;DR on his take is that there are organizational and cultural issues that prevent Google from focusing their research efforts in the way that is necessary for what he calls "big swings," like training GPT-3.

In regards to your second question, Google's reputation in ML is definitely not hype. Purely on the research side, Google has been behind some of the most important papers in modern ML, particularly around language model. The original Transformers paper, BERT, lots of work around neural machine translation, all of the work that DeepMind has done post-acquisition, and the list goes on. On the applied side, they also have some of the most successful/widely-adopted ML-powered products on the market (think RankBrain/anything involving a recommendation engine, Translate, Maps, a ton of functionality in Gmail, etc).

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

#58

Post by PartiallyTyped »

thesimp wrote: post #37

I'm actually shocked that a company that has spent 25 years on finetuning search results for any random question people ask in the searchbox does not have a good, clean, dataset to train an LLM on. Maybe this is the time to get out the old Encyclopedia Britannica CD and use that for training input.

You may find this illuminating. The google prior to 2019 isn’t the google of today.

https://www.wheresyoured.at/the-men-who-killed-google/

Edit: there was also a discussion on HN about that article.

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

#59

Post by sebastiansm »

pilooch wrote: post #15

So Google hasn't used an LLM to generate and test weird queries ? This is not putting the bar very high for the whole industry... There'd be so much to gain from a clean deployment... Either it hard, either it is a rush. As a machine learnist, I believe it's actually impossible, by design of the autoregressive LLM. This race may we'll be partially to the bottom.

Google is working hard to be the next Boeing.

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

#60

Post by jsemrau »

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…

This is statistics though. Edge cases are nothing new and risk management concepts have evolved around fat tails and anomalies for decades. Therefore the statement is as naive as writing a trading agent that is 100% correct. In my opinion, this error shows lack of understanding responsible scaling architectures. If this would be their first screw up I wouldn't mind, but Google just showed us a group of diverse Nazis. If there is a need for consumer protection for online services, it is exactly stuff like this. ISO 42001 lays out in great detail that AI systems need to be tested before they are rolled out to the public. The lack of understanding of AI risk management is apparent.
Post reply on HN