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

361–370 of 391 posts

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

#361

Post by mark_l_watson »

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 am not really disagreeing with you, but just wanted to add:

Probably the secret sauce to using LLMs for search and other forms of information retrieval is first taking a query and use traditional lookup to find likely information. Then aggregating this data, provide the aggregated data as context, append the query, and pass it to the LLM. Probably some post processing is a good idea also.

I used to do this with squad LLMs.

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

#362

Post by nmca »

sho_hn wrote: post #245
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'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…

GopherCite, Sparrow, WebGPT and most likely the new bing (WebGPT2) all first find sources then generate output, so you're just mistaken. The first two specifically get folks check whether the selected source snippet supports the answer.

Source: worked on GopherCite.

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

#363

Post by thefurdrake »

MaxDPS wrote: post #340
isoprophlex wrote: post #274

Earlier quoted context omitted.

When was the last time Google delivered something that was actually good/lasted longer than an afternoon? They have had search, maps and gmail since forever. To me, it seems they don't know how to innovate anymore, let alone deliver. It'll be the story of YouTube Shorts, (stadia, ... you name it) over and over, until they fade into obscurity.

Android? Chromebook?

Android: 2003 Chromebook: 2011

I feel like there have to be better examples.

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

#364

Post by pellucide »

jaggs wrote: post #91

https://www.youtube.com/watch?v=npV4Kix7Td0 If this was a startup, they would be dead in the water. Nothing of substance, awful sound, 9:45 mins in they don't have a key piece of live demo equipment (the phone)/ As someone else said, it's like they asked a couple of interns to rush something out in 24 hours. No wonder the stock price is tanking. Awful display from Google.

I am in the minority here. But for me, google's presentations were always like this. The presentation always seemed so rehearsed with artificial excitement.

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

#365

Post by JustLurking2022 »

htrp wrote: post #21
guardiangod wrote: post #9

I hope Google is not pulling a Blackberry moment. I remember when the iPhone was first announced, BB scrambled to put out press releases saying they have competing products in development. 2 years later the Bold(EDIT: Storm) was released and everyone realized it's just a hastily designed iPhone clone. That marks the end of BB's dominance. >Used to work for BB/RIM

> 2 years later the Bold was released and everyone realized it's just a hastily designed iPhone clone. That marks the end of BB's dominance. I feel like Google is going to bleed credibility for every month that they don't have a lamda/llm enabled search integrated into the home page. You can talk all you want about the foundational models and how advanced they are, but the search product itself will require a ton of…

Strongly disagree - they will lose credibility if they put out something half-baked. Yahoo existed long before Google search - they won because the experience was so much better that people started calling "Googling" rather than searching.

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

#366

Post by gardenhedge »

jaggs wrote: post #91

https://www.youtube.com/watch?v=npV4Kix7Td0 If this was a startup, they would be dead in the water. Nothing of substance, awful sound, 9:45 mins in they don't have a key piece of live demo equipment (the phone)/ As someone else said, it's like they asked a couple of interns to rush something out in 24 hours. No wonder the stock price is tanking. Awful display from Google.

I just skimmed through the video and didn't see anything too terrible. To be honest, I don't care they forgot a phone. I'd rather see mistakes like that than over rehearsed presentations.

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

#367

Post by gardenhedge »

TrevorJ wrote: post #180
Wowfunhappy wrote: post #135

Earlier quoted context omitted.

> I feel like Google is going to bleed credibility for every month that they don't have a lamda/llm enabled search integrated into the home page. You can talk all you want about the foundational models and how advanced they are, but the search product itself will require a ton of fine-tuning on real-world training data that they aren't collecting yet. I love ChatGPT as a source of inspiration, but I don't think I'd w…

Google already has an issue with accuracy - the snippets they choose are often not very correct.

It's incredibly frustrating. They're often completely wrong.

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

#368

Post by MaxDPS »

thefurdrake wrote: post #363
MaxDPS wrote: post #340

Earlier quoted context omitted.

Android? Chromebook?

Android: 2003 Chromebook: 2011 I feel like there have to be better examples.

Chromecast, Android Auto, Wear OS

I'd personally give Google Colab an honorable mention since it seems like the go to notebook for sharing ML research/projects.

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

#369

Post by 908B64B197 »

guardiangod wrote: post #338
908B64B197 wrote: post #120

Earlier quoted context omitted.

I remember the Storm. They released it one month or so before or after Apple released the iPhone 3G (and the App Store). I remember thinking it was a prototype when someone showed a Storm to me. Scrolling was janky, the whole phone was sluggish and slow and the keyboard barely worked. Blackberry couldn't even figure out how to get Wifi on the thing, so it was stuck using cellular while my iPhone was just streaming vi…

RIM was a hardware company, not a software company. Hence a phone is a phone, and should not be a camera (1 of the CEOs said it?) The BBOS was not designed to be a responsive, touch-screen OS. It had its root in pagers (hence the jankiness). Thru the grapevines what I heard was that the OS team said 'we can duplicate iOS' functionality with the old OS in this amount of time.' and the higher ups rolled with it. By the…

> RIM was a hardware company, not a software company. Hence a phone is a phone, and should not be a camera (1 of the CEOs said it?)

So was Apple. Other than their brief attempt at licensing MacOS, they were and still are a hardware company at heart.

> I don't have any specific insight into the Wifi stack but most BBs up to that point have in-house designed cellular modems.

They were maintaining their own modem implementations? That's quite a feat.

> As for talent wise, shrug. You can't blame people for preferring Cupertino over Waterloo-Kitchener. Although I would say that over 1/5 of the employees at RIM's campus were co-op students.

Was compensation even remotely comparable? Having 20% of the company be interns is... rather unusual.

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

#370

Post by uoaei »

jakeinspace wrote: post #277
uoaei wrote: post #267

Earlier quoted context omitted.

Any numerical system operates only on numerical, not logical/grammatical/whatever constraints. Wrongness is built into any system based solely on neural components by the simple fact that rightness is unverifiable from within these systems' own "logical" (read: associational) apparati. I share your confusion regarding the grave dearth of incredulity in the reception of these technologies.

I guess you can say the human brain operates only on the data it gets from our ear drums, optic nerve, etc. Might as well call that numerical input, it’s certainly not grammatical or logical, all that is internal and arises after many layers of neurons.

I agree: the human (or any organic) brain is no more "logical" than any other numerical system. We obscure this fact (mostly) successfully by virtue of the statistical regularities with which we are imprinted as a result of our experiences. This gives the appearance of logical rigor, but this rigor is only illusory.
Post reply on HN