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

Google’s not-so-secret new OS

techspecs.blog

Post reply on HN
Print view View original

371–380 of 561 posts

Re: Google’s not-so-secret new OS

#371

Post by flukus »

taneq wrote: post #351
michaelmrose wrote: post #256

Earlier quoted context omitted.

This outlook is why lots of people have to go to work and deal with absurdly slow trash software written to run on absurdly fast hardware.

No. I'm not saying terribly designed, lazily coded enterprise apps are ok, I'm saying if Java can be fast on a $200 mobile phone, it can be fast on a desktop.

> I'm saying if Java can be fast on a $200 mobile phone, it can be fast on a desktop.

It's not fast on a $200 mobile phone though. It's still pathetically slow.

Re: Google’s not-so-secret new OS

#372

Post by fauigerzigerk »

bitL wrote: post #333
fauigerzigerk wrote: post #323

Earlier quoted context omitted.

Sounds like something that could be used for customizing a CMS a little bit, but software development is something very different from what you're describing. Developing software requires understanding of completely open ended natural language. NLP is nowhere near that level of AI and doubt that it will be in the next 30 years.

Yes, but when you talk to regular people and their needs for web pages, apps, they are often either very trivial or unbearably complex. You can potentially automate away those very trivial with the current state of ML already, and there is a bulk of money there that goes to a lot of independent developers and smaller companies. And once you have such a system built, you can extend it as new advances in ML/GPU come, a…

>Yes, but when you talk to regular people and their needs for web pages, apps, they are often either very trivial or unbearably complex

Mostly they are unbearably vague and based on tons of false assumptions about how things work. Separating the trivial parts of a user request from the unbearably complex parts is itself often unbearably complex. It requires a conversation with the user to make it clear what is simple or what could be done instead to make it simpler.

The examples for trivial user needs that you have given are all within the realm of what we now use WISIWIG editors for. Not even that is working well. The problem is that you can't layout a page without understanding how the layout interacts with the meaning of the content on the page.

The logic capabilities of current ML systems are terrible. It's like, great, we have learned to sort numbers almost always correctly unless the numbers are greater than 100000!

Even in areas where AI has advanced a lot recently, like image recognition, the results are often very poor. I recently uploaded an image I took of a squirrel sitting on tree branch eating an apple to one of the best AIs (it was Microsoft's winning entry to some ImageNet competition).

It labelled my image "tree, grass", because the squirrel is rather small compared to the rest of the picture. Any child would have known right away why that picture was taken. The tree and the grass were visually dominant but completely unremarkable.

Re: Google’s not-so-secret new OS

#373

Post by TeMPOraL »

pjmlp wrote: post #339
TeMPOraL wrote: post #303

Earlier quoted context omitted.

Bump it to 3-4 years then. Bloated manufacturer updates combine with bloat in most popular applications and the regular web bloat to make the phone unusably slow after few years. Between that and fragility of smartphones (mechanical damage, water damage), most people are bound to replace theirs every few years anyway.

> Bump it to 3-4 years then. Bloated manufacturer updates combine with bloat in most popular applications and the regular web bloat to make the phone unusably slow after few years. My S3 is 4 years old now, and it is works perfectly fine. When it dies, I will most likely adopt one of my Lumia devices as main one, or will buy a 2nd hand Android device, instead of giving money to support bad OEMs

My SO's S3 is of more-less the same age and it's so slow that it's barely usable now. Still can't track down why - she is not a power user, she wasn't installing apps beyond the few things I installed her and the OS updates. My old S4, currently used by my brother, suffered the same fate, being slow even after a factory reset. I wonder where this comes from?

Re: Google’s not-so-secret new OS

#374

Post by flukus »

regularfry wrote: post #283
babesh wrote: post #164

Earlier quoted context omitted.

A product I worked on was a victim of Microsoft changing preferred APIs. Their competing product which the prior year had less market share somehow sported the new UI instantly when the UI became publicly available. Monopolistic behavior if you ask me.

It's why I got out of the MS ecosystem. I bought quite heavily into WPF for the Vista launch, and it was obvious that was a dead end within a couple of years. Not only that, but I could see exactly the same thing happening to all the other neat stuff I was planning to look into at the same time. That was not the way to encourage my long-term membership of the Visual Studio clan.

> and it was obvious that was a dead end within a couple of years

As the other reply probably indicated, it's still not apparent to .net developers. MS really dropped the ball with providing a clear path for desktop development.

Re: Google’s not-so-secret new OS

#376

Post by SiVal »

chriswarbo wrote: post #346
TeMPOraL wrote: post #303

Earlier quoted context omitted.

Bump it to 3-4 years then. Bloated manufacturer updates combine with bloat in most popular applications and the regular web bloat to make the phone unusably slow after few years. Between that and fragility of smartphones (mechanical damage, water damage), most people are bound to replace theirs every few years anyway.

> Bump it to 3-4 years then My smartphone's getting on for 9 years old; I don't think I'm a key demographic ;)

But you may be the leading edge of what could easily become a "key demographic": people who don't upgrade for many years. It's possible that "things I could do with a new phone that I can't with my existing phone" will be worth less than the cost of a new phone for an increasing number of people. Or not. I don't think anyone knows for sure how that will evolve over the next decade or so.

Re: Google’s not-so-secret new OS

#377

Post by throwawaydbfif »

bertrand-caron wrote: post #307
throwawaydbfif wrote: post #228

Earlier quoted context omitted.

I've used a ton of languages over the years and vastly prefer Java type syntax when working on larger projects. The forced organization tends to lead towards some level of mandatory code clarity. Something greatly lacking in Js land. OO is a bad word these days and functional is all the rage, even though functional languages were largely superceded by OO languages eons ago for many reasons people are slowly redicover…

Could you elaborate on how "functional languages were largely superceded by OO languages eons ago for many reasons people are slowly redicovering" ?

Erlang and common lisp have been around for a long time, and functional programming is nothing new. The reality is that most business problems map conceptually to communication between objects, and that IDE's which greatly help developer productivity work a lot better with objects.

Functional programming has origins in lambda calculus and academia because mathematical problems map more easily from pure math to functional programming. It's really popular in the circles where it's more useful/easier than OO.

Honestly I don't think the people 20 years ago chose OO for most business languages over functional out of ignorance. They had a choice and decided that OO was better for business problem solving languages like Java even though a large majority of programmers from that era were math majors and familiar with functional syntax.

I feel like we're in one of those cycles where a large number of a previous generation have retired and it's time to learn some of these lessons all over again.

Notice how many wood commercial buildings have been going up in the last 15-20 years? A lot, and just long enough after everyone involved in all the great city fires of WW2 to be too dead to object.

Re: Google’s not-so-secret new OS

#378

Post by majewsky »

ianai wrote: post #185
felixge wrote: post #162

Earlier quoted context omitted.

> In two years, your average consumer probably doesn't care if their device is still receiving software updates. They'll jump onto a new phone plan, with a fresh, cool new mobile, with a better screen, newer software (features!), and a refreshed battery. This might be true for the US, where 75% of subscribers are on post-paid (contracts). It's not true for the rest of the world. * Europe: * Rest of the world: I'd als…

I think the "post paid" connection leading to a 2 year lifecycle is suspect. There's a big after market repair industry in the US. Many people have a singular cell phone as their internet device - and it's often ancient by IT terms. 2+ year old hardware needs to be getting software/OS upgrades.

I favor the proposal of requiring a prominent "Best Before" date for new devices, indicating how long the manufacturer will guarantee the availability of security updates.

Re: Google’s not-so-secret new OS

#379

Post by flukus »

bashtoni wrote: post #259
bitL wrote: post #247

Seems like Free Software that propelled early Internet pioneers served its purpose and those companies are turning their backs on it - first with Apple, GCC->LLVM, now with Google, Linux->Fuchsia :( I am getting afraid of another dark age on the horizon... I guess it's going to be inevitable as 90% of SW developers will find themselves redundant when inferring AI capable of composing code blocks and listening to/read…

I think Google is probably the most Free Software friendly of the the new big three (Amazon, Google, Microsoft). They haven't disappointed with Fuschia which appears to be entirely copyleft: https://fuchsia.googlesource.com/magenta/+/master/LICENSE https://fuchsia.googlesource.com/fonts/+/master/LICENSE

Of the big three MS has definitely been the best lately, unlike the other too they've embraced open community development, not just open source. I still wouldn't trust them not to pull a bait and switch, but they've been better than the other two.

Re: Google’s not-so-secret new OS

#380

Post by bitL »

fauigerzigerk wrote: post #372
bitL wrote: post #333

Earlier quoted context omitted.

Yes, but when you talk to regular people and their needs for web pages, apps, they are often either very trivial or unbearably complex. You can potentially automate away those very trivial with the current state of ML already, and there is a bulk of money there that goes to a lot of independent developers and smaller companies. And once you have such a system built, you can extend it as new advances in ML/GPU come, a…

>Yes, but when you talk to regular people and their needs for web pages, apps, they are often either very trivial or unbearably complex Mostly they are unbearably vague and based on tons of false assumptions about how things work. Separating the trivial parts of a user request from the unbearably complex parts is itself often unbearably complex. It requires a conversation with the user to make it clear what is simple…

Just imagine that you can interactively by voice or by touch tell AI what/how to adjust stuff and it will use it to improve itself for your future similar tasks. Now project there will be 1,000,000 users like that, telling app what exactly did they mean and pointing to proper places in the app. So exactly this will be the conversation you desire, you'd directly tell your app builder what you want, and if it is not doing what you like, you either show it to builder by simple gestures or rely on some other user going through the same problem before you and app builder taping onto that knowledge. Obviously, first for simpler web or mobile apps. This sounded like sci-fi just a decade ago, but we now have means to do simple app builders like that.

ML by itself is incapable of inference, hence you need some guiding meta-programming framework that could integrate partial ML results from submodules you prepare.

As for squirrel example, it was probably one of "under threshold" classifications of ResNet, i.e. tree was 95%, grass was 90%, but squirrel was 79%, so it got cut out of what was presented back to you. Mind you, this area went from "retarded" in 2011 to "better than human in many cases" in 2016. I know there are many low-hanging fruits and plenty of problems will still be out of reach, but some are getting approachable soon, especially if you have 1M ML capable machines at your disposal.

Post reply on HN