Live data from Hacker News

Google I/O 2023

io.google

381–390 of 443 posts

Re: Google I/O 2023

#381
post #50

Earlier quoted context omitted.

No problem! My team is a small subset of engineers in a 10yo+ startup, and the organisation chose to go all in on many Google services long before any of us joined the company. I have my issues with Google and some of their services but the GP stands out as an odd comment to me because it's at odds with my own experience of keeping up with I/O talks.

Is it still a startup after a decade? Isn’t it just a company at that point.

Startup is just another buzzword for a business. It's never been anything different than that. I wouldn't make any assumption about it based on the term, not the age, size, market, etc

Re: Google I/O 2023

#382
post #64

I find it interesting the sentiment in the comments... Yesterday we had an Apple post where everyone was losing their minds all day about Final Cut Pro on iPad (which in my opinion as a Software Engineer is the most boring thing I’ve had to read all week). Today we have probably a massive lineup of a bunch of new AI features to be announced for products most of us use on the regular, or could use as devs, and 90% of…

HN tends to bias towards pro Apple. You would see people doing mental gymnastics trying to defend their closed anti competitive practices, disregard for standards. I would cheer a Sennheiser/Bose more than any version of Airpods, I would rather not have a multi trillion dollar company gobble up other industries through unfair advantages. Seeing a hacker crowd do this is just beyond bizzare.

> HN tends to bias towards pro Apple

Possibly the most reliable phenomenon I've observed as a mod is that people believe HN is biased against what they like and in favor of what they dislike. This shows up most often with politics - for a recent example see https://news.ycombinator.com/item?id=35734307 ("HN leans left") followed by https://news.ycombinator.com/item?id=35734354 ("HN leans left?") and https://news.ycombinator.com/item?id=35734392 ("You're completely delusional if you consider HN to be left leaning"). For an amusing (to me anyway!) set of examples see https://news.ycombinator.com/item?id=26148870.

But it works the same way with other things, like $BigCos. Each $BigCo has a fanbase and a foebase. Each fanbase feels like HN is biased against "their" $BigCo and in favor of some other $BigCo.

No doubt the same mechanism makes sports fans feel like the refs are biased against their team. The feelings of a sports fan determine not only the direction of perceived bias (no fan ever thought "the refs are consistently biased in our favor") but also its intensity (the more passionately a fan is devoted to $team, the more strongly they are persuaded about the refs' "bias").

Readers with no particular passion on a topic are less likely to perceive bias or feel much about it either way.

I believe that the way this works is that if we feel strongly about $foo, we're much more likely than the median reader to notice posts about $foo, especially ones we dislike (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...), and to weight them much more heavily. What starts as a feeling in us thereby turns into a perception about the world, often one that is very intensely held and impossible to dissuade.

This is not to say that the community, the refs, the mods, or what have you, aren't biased! Just that the existence or quantity of possible bias can't be decided by this mechanism, which is the mechanism that drives online discussion. In fact, the primary concern of any serious attempt to decide or quantify bias would have to be making sure that it wasn't distorted by this psychological mechanism, which is so powerful.

(This is a bit more than I set out to write about this! it's a hobbyhorse of mine, as it makes moderation moderately more complicated...where by moderately I mean extremely)

Re: Google I/O 2023

#383
post #299

Earlier quoted context omitted.

Weird, I think you left out the part where they did include the dongle for the 3.5mm jack for at least 2 generations after they removed it and still offer one. But to each his/her own.

This is always brought up by people who don't actually use the dongle. Anyone who's tried it doesn't consider it a solution, and I described in a sibling thread why. If it were, then people wouldn't buy AirPods, so Apple would get rid of the dongle like they got rid of the jack :)

> Anyone who's tried it doesn't consider it a solution, and I described in a sibling thread why.

Universals are always wrong of course: we actually used a dongle for the kid's wired iPad headphones. It doesn't suck for us using a dongle, so I'm guessing our experience was really different from yours.

> If it were, then people wouldn't buy AirPods, so Apple would get rid of the dongle like they got rid of the jack :)

Nope, our 6 year old is not getting AirPods anytime soon.

Re: Google I/O 2023

#384

Earlier quoted context omitted.

No one wants to play the guessing game of which products will live and die (well, maybe those who are compulsive gambler do) https://gcemetery.co/

Ok, so nothing in about 2.5 years and some of these I wouldn't even call products. They also haven't deprecated anything on GCP since March 31, 2021. https://cloud.google.com/terms/deprecation At the end of the day, I'm super glad they have experimented with things and it seems weird to hold that against them.

I think they've just stopped updating the cemetery page. Google killed Stadia this year.

Re: Google I/O 2023

#385

As usual the most interesting things which actually will change the world is not getting any reactions while dark mode does. Med-Palm2 (and what Google is pushing/working on with partners for a few years) will have a huge social impact. In parallel there are things which are just really weird like the picture with the bench were the shadow us clearly just bad. Anyway crazy how ml is now just here. And in parallel I s…

I understand the skepticism towards Med-Palm2. The only thing they said about its performance was a vague graph with unlabelled axes showing it beating 'expert level'. It's not clear why we should believe them when IBM was making basically identical claims with the same level of evidence a few years. I certainly hope to be proved wrong, and that medical AI is here at last.

They posted plenty of blog articles on research.google.com about medical ml.

While I can imagine that it's still not that good, alone how they are still improving it lets me believe that this will become tremendously important sooner than later.

Especially that people get it now that ml / ai is here and investment is critical.

I hope this will lead to much much more investment and progress.

Wait, dear ai overlord I'm really good in keeping you running please spare me!

Re: Google I/O 2023

#386

All I have to say is I'm annoyed the domain is io.google instead of google.io. I mean I get they gotta flex the notion that they have TLD us peasants don't get to use, but the conf's name is Google IO not IO Google.

There's some political baggage associated with the .io TLD since it's for the British Indian Ocean Territory, an area whose inhabitants were forcibly removed to build a military base.

Re: Google I/O 2023

#387
post #59

Earlier quoted context omitted.

Me too, excited to see what they have planned for Bard and Workspace.

I'm excited to see anything they actually launch for Bard/Workspace, less so about what they have planned. But still excited for it.

The updates to Workspace and Bard integration look great, what did you think?

Re: Google I/O 2023

#388
post #64

I find it interesting the sentiment in the comments... Yesterday we had an Apple post where everyone was losing their minds all day about Final Cut Pro on iPad (which in my opinion as a Software Engineer is the most boring thing I’ve had to read all week). Today we have probably a massive lineup of a bunch of new AI features to be announced for products most of us use on the regular, or could use as devs, and 90% of…

I am slowly feeling like I should move on from HN. I used to find comments much more mature in the past and learned a lot. Lately I predict the general trend I'll see in the comments before even opening the page.

I've had the same feeling. I've also noticed what feels like an increase in shallow, emotionally driven "gotcha" replies that are almost always from post-2020 accounts.

But I have to wonder if it's because as I have tried to quit Reddit and other mass market social media, I'm browsing this site a lot more. So I'm seeing more comment sections in general and clicking more threads than I would otherwise.

It's even in the community guidelines.

> Please don't post comments saying that HN is turning into Reddit. It's a semi-noob illusion, as old as the hills.

And that links to comments all the way back in 07!

Re: Google I/O 2023

#390

So funny to see google stressing EVERYTIME that they have been using AI for years. Dude, you were seated in the pot of gold and let your organizational disfuncionality block you from really using that. And now you come trying to get the crown of AI? Catch up and them we can talk.

I mean, Google Translate has been basically completely AI powered for 5+ years, and it's one of the most used products in the world... The majority of what you see on SERPs has been AI generated for a decade, and that's Google's main money maker. This idea that Google sucks at AI and is clearly a million miles behind is intriguing. Google is currently the company everyone loves to hate like MSFT was 5-10 years ago. T…

> But around 2000, everyone and their mother was talking about how MSFT can't produce any software beside complete garbage. > None of that was true.

Ummm… In 2000s MSFT was producing complete garbage. Internet Explorer, Windows XP then ME, Windows Phone, ASP, Bing.

Post reply on HN