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

Google has sent internet into 'spiral of decline', claims DeepMind co-founder

telegraph.co.uk

Post reply on HN
Print view View original

241–250 of 390 posts

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#241

Post by brucethemoose2 »

donatj wrote: post #52

I wrote a not-very-coherent in retrospect article about 12 years ago while working for an SEO firm about how SEO was killing the internet. The basis being companies were creating pages and blog posts specifically to target keywords without actually having coherent things to say on the topic. Moreover people within Google like Matt Cutts seemed to be at the time if not outright encouraging it, at the very least giving…

Do you have a link?

Twelve years ago, this wasn't even a thought in the back of my head, and I bet those in the SEO industry never thought Google would let things deteriorate to such an extreme.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#242

Post by wuming2 »

I never understood people writing as there were, or are, no alternatives. For ad professionals perhaps or for content providers. Of any quality. But as a consumer all I needed to leave Goog was to try a few other search engines and settle on the one returning results closest to what I wanted. Then tapped Settings > Safari > Search Engine and made my choice. Years ago now.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#243

Post by moritzwarhier »

EForEndeavour wrote: post #222
Am4TIfIsER0ppos wrote: post #192

Earlier quoted context omitted.

> git reset --hard HEAD~2 Stupid computer. That does 2 commits. > git reflog expire --expire-unreachable=now For the love of god why? The reflog is there to save your bacon. I guess because they are still reachable from the remote's reference so they won't completely vanish just yet. > git push -f Now they're gone unless you can find the hashes somewhere else. The objects might still be present. > I apologize that tw…

Nah, they likely didn't bother trying to ask GPT-4 and imagined hypothetical answers which could lead beginners to shoot themselves in the foot. ChatGPT Plus with a short custom instruction to answer in steps, respond tersely, ask clarifying questions where needed, and make recommendations wherever possible replied to "how to reverse a pushed commit in git" with: --- To reverse a pushed commit in Git, you can either…

You are correct! I was fantasizing how a bad GPT answer could look like, similar to the errors that were in the answers to my prompts to GPT-3.5 for imagemagick. These prompts where precise and in good faith, with some rounds of clarification.

I am sure newer GPT versions will do better, and I think for a tool as ubiquitous as git the answers from ChatGPT would have been perfectly fine too.

Just wanted to illustrate why I wouldn't mske it a habit to ask it factual questions.

The "lost two commits" answer was meant as whimsy joke. More realistic scenarios are more boring of course.

E.g. memorizing wrong explanations for solutions that happen to work, or accepting wrong answers as fact when it's inconsequential for the short term acceptance of the answer.

The same thing easily happens with Google's knowledge snippets they extract from search results.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#244

Post by jdub »

dalbasal wrote: post #121

"Search results plagued with ‘clickbait’ would have been an understatement 20 years ago. The parts of the open web you access from Google/search (most of it) is almost entirely made-for-google' ranking algorithims. The situation is worse than it is on closed social media, including Google's YouTube. Professional YouTubers are also making content for rankings above all, but the fact that it's their own face/voice keep…

Vint Cerf

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#245

Post by bad_username »

NL807 wrote: post #199
dalbasal wrote: post #121

"Search results plagued with ‘clickbait’ would have been an understatement 20 years ago. The parts of the open web you access from Google/search (most of it) is almost entirely made-for-google' ranking algorithims. The situation is worse than it is on closed social media, including Google's YouTube. Professional YouTubers are also making content for rankings above all, but the fact that it's their own face/voice keep…

I'm rather shocked how the quality of their search results has tanked over the years. I used to be able to find decent academic content, or programming related topics quite consistently 5 years ago. Now I'm getting only a few vague results either from stackoverflow, math stackexchange, reddit, or apple, and the rest is just utter garbage. Yandex, of all search engines provide consistently better results. Fortunately,…

The Russian search engine pops up in the comments whenever Google search quality is discussed here. I am sure this is non-organic astroturfing.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#247

Post by brucethemoose2 »

xkcd1963 wrote: post #53

It's not like we can't use the internet without google.

People will largely fall back to other "lazy" search methods (like just asking ChatGPT (and getting a confidant wrong answer) or searching TikTok or YouTube (and getting a influencer with bad information)) rather than some more rigorous search method.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#248

Post by moritzwarhier »

manmal wrote: post #119
moritzwarhier wrote: post #93

Earlier quoted context omitted.

> But with the replication crisis in many sciences and political meddling abound, I‘d be wary of any kind of critical information, especially on Wikipedia. Of course, transformer models have the same issues I think this is a bad comparison. The issues OP described are not limited to tainted training data. Being critical of Wikipedia and critical of science is not equal to treating text as a black box looking for the…

Can you point to where in my comment I asserted that hallucinations are equal to other sources of errors, or even compared them? Because I tend to agree with you. I tried to point out that info found externally always has to be cross-checked, no matter where it originates.

I meant the sentence I quoted, or the part

> [...] especially on Wikipedia. Of course, transformer models have the same issues

I agree that one should be critical of all sources, including Wikipedia. I don't agree that GPT is an information source comparable to Wikipedia or scientific studies. Both can be wrong, biased or incorrect though.

In other words, I would never consider the usage of any LLM worth to cross-check information found elsewhere, as opposed to e.g. Wikipedia.

Which does not mean that I consider the info compiled on Wikipedia always as trsutworthy. Then again, I wouldn't use an LLM to cross-check Wikipedia.

They're not equals.

Sorry if I misunderstood anything in your comment.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#249

Post by brucethemoose2 »

ramraj07 wrote: post #158
Jensson wrote: post #140

Earlier quoted context omitted.

How do I publish data to an LLM? I don't see how it solves the issues the open web solved, so it doesn't seem like a good replacement to a search engine.

As I said we are still in the Lycos era of this paradigm. But I’m not sure what you mean, the LLM is the index of the entire internet, so not sure why it’s hard to imagine it becoming a much smarter answering machine (answers being either the actual answer or a link to a website that’s actually useful).

> LLM is the index of the entire internet

It is certainly not an index of the entire internet. Web scraping is a tool to make LLMs "smarter" and better at speaking, but they are not a web archive, not even a reasonably lossy one.

Honestly they are a poor substitute for Google, even when augmented with a vectorDB or (hypothetically) some kind of confidence metric, when the information is not immediately verifiable.

Re: Google has sent internet into 'spiral of decline', claims DeepMind co-founder

#250

Post by TerrifiedMouse »

mrtksn wrote: post #238
donatj wrote: post #52

I wrote a not-very-coherent in retrospect article about 12 years ago while working for an SEO firm about how SEO was killing the internet. The basis being companies were creating pages and blog posts specifically to target keywords without actually having coherent things to say on the topic. Moreover people within Google like Matt Cutts seemed to be at the time if not outright encouraging it, at the very least giving…

> I have to imagine the growth of large language models is only going to amplify this practice dramatically. Funny enough, LLMs are actually for the very first time challenging Google’s utility as a knowledge engine. I see more and more Non-techie people asking ChatGPT about receptors or locations instead of Google. Also, if there’s anything that I need to research and it’s not a general knowledge stuff but something…

Will it last though. How soon until LLMs get gamed as well - probably directly, e.g. looking for a product that can do X, SEOs battle it out to make sure a specific product gets suggested by the LLM.

There are many problems with LLMs as knowledge engines,

- as mentioned they can be gamed just like Google

- they hallucinate, adding another possible source of errors to the results

- they kill off the websites they scrap training data from making it unsustainable long term

Post reply on HN