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

Google's live from Paris event private/deleted immediately

news.ycombinator.com

Post reply on HN
Print view View original

381–390 of 391 posts

Re: Google's live from Paris event private/deleted immediately

#381

Post by indymike »

acdha wrote: post #189
indymike wrote: post #141

Earlier quoted context omitted.

This isn't true at all. The G1 (first widely marketed android phone) was superior in all ways but one: Marketing. The iPhone wasn't really competitive on the hardware front until the 3g which finally added GPS and a fast cellular modem.

The iPhone 3G was released before the G1, and period reviews especially noted how the G1 didn't measure up. In general, if something non-free is wildly successful the explanation is more than “marketing” and that certainly sounds like it's the case: > The HTC hardware and Android OS that powers it lack the polish and depth of even the iPhone 1.0 in most respects. > My first-generation iPhone with iPhone OS 2.1 feels…

I was wrong. The 3GS was the first model with GPS, which came about five months after the G1. The G1's touchscreen was very comparable to the Apple devices I was building software for at the time - I think the multitouch capability was there, and really the only use was pinch to zoom... which I seem to recall working on the G1 after an OS update. Regardless, the web browser part was rather the opposite - the G1 with it's weird trackball could be used to browse desktop sites and at the time, most websites were not responsive. Also, there were places where android used long-press... which Apple added later. Regardless, I was fascinated at the time how Apple was able to outmarket everyone with iPhone... and equally fascinated with how quickly the army of Android vendors materialized. Fun times.

Re: Google's live from Paris event private/deleted immediately

#382

Post by acdha »

indymike wrote: post #381
acdha wrote: post #189

Earlier quoted context omitted.

The iPhone 3G was released before the G1, and period reviews especially noted how the G1 didn't measure up. In general, if something non-free is wildly successful the explanation is more than “marketing” and that certainly sounds like it's the case: > The HTC hardware and Android OS that powers it lack the polish and depth of even the iPhone 1.0 in most respects. > My first-generation iPhone with iPhone OS 2.1 feels…

I was wrong. The 3GS was the first model with GPS, which came about five months after the G1. The G1's touchscreen was very comparable to the Apple devices I was building software for at the time - I think the multitouch capability was there, and really the only use was pinch to zoom... which I seem to recall working on the G1 after an OS update. Regardless, the web browser part was rather the opposite - the G1 with…

A lot of us were also there at the time. It was considerably more than marketing, and a lot of companies said they had iPhone killers which promptly failed in the market. It took a solid half decade for the competition to really firm up because there were a lot of little details which required time to get right.

It was definitely eye-opening to see people’s usage habits shift so quickly. I know multiple large organizations who completely reapproached how they developed internal applications because the executives no longer accepted IT’s unwillingness to support mobile access. I’d previously thought that would never happen.

Amusingly, the one which had a damaging public breach did not have it due to allowing phones access as some had predicted but when one of IT’s managed laptops was stolen out of someone’s car and, unlike an iPhone, it didn’t have storage encryption enabled.

Re: Google's live from Paris event private/deleted immediately

#383

Post by khmii »

throwawayapples wrote: post #297
visarga wrote: post #250

Earlier quoted context omitted.

Search "What is the world record for crossing the English Channel entirely on foot" and see precision Zero. None of the responses match the query. It should say - not found. Or actually find that people attempted to illegally cross through the Channel Tunnel. Usually when I search for error messages for a specific version of software it will respond with other versions. That is useless. That's not what 24 years worth…

asking ChatGPT gives very interesting results (didn't check accuracy): > What is the world record for crossing the English Channel entirely on foot As of my training data, the world record for crossing the English Channel entirely on foot is held by Christof Wandratsch, who completed the swim in 8 hours and 58 minutes on August 25, 2018. > Swimming isn't on foot My apologies, you are correct. The English Channel is t…

It just told me a different person too:

"The world record for crossing the English Channel entirely on foot (without the aid of any swimming or flotation devices) is held by Belgian endurance swimmer and long-distance runner, Kevin Swartz. He completed the 35-mile (56 km) journey from England to France in just 13 hours and 54 minutes on August 17, 2019."

chatGPT really is a brilliant compulsive liar.

Re: Google's live from Paris event private/deleted immediately

#384

Post by hn_throwaway_99 »

PaulDavisThe1st wrote: post #243

You know, whatever the merits (and it sounds as if there were very few) of Google's presentation (or Pinchai's recent non-statement) might be, I would have expected the commentariat here at HN to display way more skepticism about language models as a replacement for search. If there's a crowd anywhere in the world large enough to leave 200+ comments on a post like this, you'd think thatg the one at HN would understan…

> I would have expected the commentariat here at HN to display way more skepticism about language models as a replacement for search.

Because plenty of people don't see it necessarily as a replacement for search, but as a useful adjunct.

More to the point though, why do people always want to define the "HN commentariat" as some monolithic bunch? I've seen tons of comments and articles on the front page pointing out many issues with LLMs, from attribution, to confident bullshitting, to bias issues. Heck, there was plenty of discussion about a big proposed reason for Google's stock crash yesterday was due to the ad Google used to highlight questions their LLM can answer gave a confidently incorrect response.

Whenever I see comments that treat the rest of the HN audience like some monolithic group, especially in the face of plenty of evidence to the contrary, I feel like, ironically, the poster is saying "Why don't you all agree with me?"

Re: Google's live from Paris event private/deleted immediately

#385

Post by vel0city »

rchaud wrote: post #316
yao420 wrote: post #261

Earlier quoted context omitted.

TikTok believe it or not. I had a 19yr old cousin tell me that she almost uses TikTok exclusively for tutorials and when I pushed back and said why/how the other teens in the family heavily agreed. I’ve since asked ~5 gen z (I don’t interact with many!) and they told me the same.

I've heard this anecdotally, but I kinda still don't believe it, as TikTok seems like an entertainment first avenue, not for more substantive learning. However, I could be completely wrong and TikTok video is the future of search. If so, where does that leave a text-only ChatGPT?

I wouldn't be too surprised. It kind of depends on the topic at hand but for instance I've learned a lot about what I know wrenching on my own motors from youtube videos.

Re: Google's live from Paris event private/deleted immediately

#386

Post by hn_throwaway_99 »

Karawebnetwork wrote: post #275
amelius wrote: post #170

Earlier quoted context omitted.

> Perplexity does it well in my opinion. Nope: https://news.ycombinator.com/item?id=34685862

I only meant about the use of footnotes links. The AI themselves are quite bad but can also be useful if you learn not to take everything at face value. One example of it failing is when I search for my own name it mixes up a bunch of people together so I end up being a dead murderer from the 70's, a house rental person, a developer and also a pianist.

Footnote links still don't solve the financial model, primarily because nobody clicks on them.

There was congressional testimony by the founder/owner of "Celebrity Net Worth" about how Google made it impossible for them to stay in business. Whenever somebody would search "How much is worth?", the answer would just show up directly on the Google results page. There was still an attribution link to Celebrity Net Worth, but nobody ever clicked on it anymore, so the result was Celebrity Net Worth had to shut down.

You can certainly argue fairly whether sites like CNW deserve to exist in the first place, but it's not hard to see how there is still a huge financial problem when ALL the ad revenue goes to the search engines and they don't even leave any of the slim scraps to the publisher sites.

Re: Google's live from Paris event private/deleted immediately

#387

Post by PaulDavisThe1st »

hn_throwaway_99 wrote: post #384
PaulDavisThe1st wrote: post #243

You know, whatever the merits (and it sounds as if there were very few) of Google's presentation (or Pinchai's recent non-statement) might be, I would have expected the commentariat here at HN to display way more skepticism about language models as a replacement for search. If there's a crowd anywhere in the world large enough to leave 200+ comments on a post like this, you'd think thatg the one at HN would understan…

> I would have expected the commentariat here at HN to display way more skepticism about language models as a replacement for search. Because plenty of people don't see it necessarily as a replacement for search, but as a useful adjunct. More to the point though, why do people always want to define the "HN commentariat" as some monolithic bunch? I've seen tons of comments and articles on the front page pointing out m…

I work really hard not to do this, and have more than once posted a message like yours.

However, my point here was that at the time of posting, the comment threads for this HN post were largely devoid of any skepticism about LM+search, and were completely dominated by a calm acceptance that it was all inevitable. And I wasn't trying to say "all of HN believes ", but rather "I would have expected there to be more voices in the comment thread expressing skepticism, because I know there are some smart people here"

Google's Bard demo certainly tweaked the situation by the end of the day.

Re: Google's live from Paris event private/deleted immediately

#388

Post by ActionHank »

ttoinou wrote: post #193
ActionHank wrote: post #55

Earlier quoted context omitted.

Having the technology or know how is still miles away from a product that is live. Google have the former, Microsoft handed over bags of cash for the latter. It's especially awkward for them because as market leader in search they can't really afford to fumble this, which anecdotally they seem to be.

Once the technology is ready and developed by others people, how much time are they really lagging behind Microsoft ? 6 months ?

Bigger cost is not time lost, but reputation impact.

People trust MS more than Google, they just don't believe that Bing is better.

Re: Google's live from Paris event private/deleted immediately

#389

Post by Jensson »

jerf wrote: post #269
sho_hn wrote: post #245

Earlier quoted context omitted.

I'd like to have more discussion about the attribution problem, specifically. We now have a couple of the players saying they're working on, or having demos. But from what I can see, all of these don't _actually_ attribute the source. They're just able to find _a_ source that fits to the output, working backward from it. Often the attributed source fits, but is actually disagreeing with the original output in specifi…

I was musing on my website about ChatGPT [1] and one of the things I noticed about it is that if you analogize it to your own brain, which has a language model in it too, we are collectively asking ChatGPT to do something we would never ask of our own language model. ChatGPT does not have the "higher level" reasoning that we're asking for, to do logic, attribution, and the other things we want. This is not necessaril…

You don't remember where you learned the things you know from? How can you trust anything you think then? Can such people get anything done at all, seems basically impossible to learn new things then since you can't know if it is better than what you thought before?

So I don't think that remembering sources is a hard problem to solve, likely making the AI work well is probably easier if it remembers sources.

Re: Google's live from Paris event private/deleted immediately

#390

Post by jerf »

Jensson wrote: post #389
jerf wrote: post #269

Earlier quoted context omitted.

I was musing on my website about ChatGPT [1] and one of the things I noticed about it is that if you analogize it to your own brain, which has a language model in it too, we are collectively asking ChatGPT to do something we would never ask of our own language model. ChatGPT does not have the "higher level" reasoning that we're asking for, to do logic, attribution, and the other things we want. This is not necessaril…

You don't remember where you learned the things you know from? How can you trust anything you think then? Can such people get anything done at all, seems basically impossible to learn new things then since you can't know if it is better than what you thought before? So I don't think that remembering sources is a hard problem to solve, likely making the AI work well is probably easier if it remembers sources.

I don't think remembering sources is an intrinsically difficult problem.

I think it's hard for language models, and indeed, after further thought, I'm pretty sure they shouldn't remember sources: https://www.jerf.org/iri/post/2023/understanding_gpt_better/

But the problem in general is probably not a particular stopper.

A bigger problem is people thinking that we should be taking language models and using them as AIs themselves, rather than seeing them as a component of an AI, which the more I think about, the more obvious it is that this is the case. Language models aren't what we want. They're a necessary step to what we do want and I don't expect them to go anywhere. But history will look back and laugh at this misguided attempt to make them be the entire AI.

Also, if you are implying you can just spew out the exact sources you learned some fact from... no you can't. You think so only because you haven't tried. Ask someone around you for five random "facts" for you to cite your source from, then try to write on a sheet of paper where you heard this fact from for the first time. Then hand the sheet of paper back to your friend and have them verify your claims... so, note, immediately you have the problem that an unverifiable claim is a fail, so "in elementary school" doesn't cut it any more than you can cite "at the library" as a source on your high school essay. I'm not looking for words that "identify" a several year span of time. You need to give me book, page, and line, or timestamps on a video, or something like that. No, you can't. Nobody can just spit out a bibliographic citation from memory for when they learned that Pluto isn't a planet. I can give you "in the news" for that, sure, but what news source was first? What is the title of that article? Who wrote it? What was the exact date? Of course not.

Post reply on HN