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

Why Apple believes it’s an AI leader–and why it says critics have it all wrong

arstechnica.com

Post reply on HN
Print view View original

61–70 of 217 posts

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#61

Post by tedeh »

jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

I'm bilingual and write approximately the same amount of messages in both languages using iOS. Autocorrect always suggests words from BOTH languages mid sentence. Like what the hell, they should be able to infer which language I'm using just from the first word.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#62

Post by toxik »

jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

The swipe keyboard is atrocious too, I simply cannot swipe the word ‘you’ with any reliability. It becomes tut, or töre. Töre. Instead of the second most common pronoun in the entire English language.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#63

Post by jolux »

IshKebab wrote: post #56
jolux wrote: post #45

Earlier quoted context omitted.

Apple’s chips are way better so whether they’re doing more locally or not they have the headroom to do a lot more. I would agree with that characterization of Google. They have a very small handful of successes and an enormous pile of failures that they’ve discarded. They give every impression of not really knowing what to do. I think Giannandrea was also referring specifically to the kinds of experiences Apple can p…

What AI failures have Google had? Google Home is much better than Siri or Alexa. The camera feature in Google Translate works really well. Their new auto subtitles on Android work really well. Searching photos for objects works really well. The only thing I can think of is that think they demoed that would call restaurants to book them for you, but that was clearly highly experimental, and it's not like Apple has don…

None of those are at risk of taking over as a revenue stream from ads on search and YouTube, is my point. You're right that it's cool stuff but right now it is being subsidized by a staggeringly more profitable business. I'm skeptical of their ability to pivot away from the dependency on search, which is good enough, obviously, but they seem to have higher aspirations of which they consistently fall short.

I really do want to see them succeed I guess I'm just overall disappointed with where they're at right now and I think they've somewhat overhyped it. I'm not a voice assistants person, I disable Siri personally.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#64

Post by jonplackett »

jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

My favourite iMessage ‘feature’ is how now and then it picks a real word and refuses to let me use it without autocorrecting it to something else.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#65

Post by mytailorisrich »

tedeh wrote: post #61
jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

I'm bilingual and write approximately the same amount of messages in both languages using iOS. Autocorrect always suggests words from BOTH languages mid sentence. Like what the hell, they should be able to infer which language I'm using just from the first word.

This also happens on Android if you have several input languages enabled.

It looks like it uses all the dictionaries so suggestions appear from all languages for every word you're typing.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#66

Post by culturestate »

strogonoff wrote: post #58
nfg wrote: post #46

Earlier quoted context omitted.

The one the drives me mad the most is when I type a word which it incorrectly autocorrects to a proper noun (which gets capitalised), I then use backspace to delete that word and type the - completely different - correct one, but it insists on upper-casing the first letter to match the original casing of the incorrect autocorrect. It really should keep track of the fact that it introduces the case-change and invalida…

I can never make iOS write “Hong Kong” when swipe-typing; tried setting up forced text replacement entry (kong->Kong) to zero effect. It is always “Hong kong”.

iOS always capitalizes Hong Kong properly for me, but I write that so much that I guess I could have trained it.

Edit: I also bought all of my iOS devices in Hong Kong so, there's that.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#67

Post by Joeri »

Veedrac wrote: post #26

Apple's AI work is decent for what it is, and their AI hardware is fine too, but so much of this article is just weird. It starts by talking about how until recently Apple wasn't doing AI work where it needed to be. Then there's the weird excerpt where he claims Google is “not known for shipping consumer experiences that are used by hundreds of millions of people.” The author then raises the legitimate point that AI…

I found the conflation between training and running models very confusing or even disingenuous. The whole thing about evaluating a model frame by frame with low latency local on the device, that is a good point, but then having that as a follow-up to why they don’t train the models with lots of data on the server ... well, I don’t really know what to make of that.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#68

Post by culturestate »

nfg wrote: post #46
jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

The one the drives me mad the most is when I type a word which it incorrectly autocorrects to a proper noun (which gets capitalised), I then use backspace to delete that word and type the - completely different - correct one, but it insists on upper-casing the first letter to match the original casing of the incorrect autocorrect. It really should keep track of the fact that it introduces the case-change and invalida…

It seems to conflate people's names with words in a sentence, e.g. "I'm crossing the Bridges." Could just be confirmation bias, but whenever I notice that problem it tends to be a name in my address book.

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#69

Post by on_and_off »

somethingwitty1 wrote: post #29
gbear605 wrote: post #24

Earlier quoted context omitted.

I’m not an Apple fankid but the predictive keyboard and speech to text work wonderfully on my iPhone and I use Siri everyday with no problems. Google’s is probably better, but the different is not a meaningful one from what I can tell.

As someone that bought the first Android phone and now has an iPhone...it is extremely noticeable how much better Android is on these fronts. But depending on the person that gap may not be important. For me, it is night and day. I felt Google had a long way to go on these fronts, so I was stunned at just how bad Apple was :(.

that's pretty much what I hear for anybody that has switched to iOS after a while on Android : they start by complaining about iOS's keyboard

Re: Why Apple believes it’s an AI leader–and why it says critics have it all wrong

#70

Post by monadic2 »

jointpdf wrote: post #34

Setting aside all the other work Apple is doing in this area, can someone explain why iMessage autocorrect is unacceptably bad (to put it diplomatically) in the era of weapons-grade language models? Its behavior is totally bizarre. It’s like an underpowered chess engine that makes flagrant blunders: capitalizing random words in the middle of a sentence (rock -> Rock), the context sensitivity of an actual rock, forcin…

I’ve found it’s much easier to disable autocorrect altogether and simply take responsibility for what you type. I mean that with no snark; that’s actually how I view it. Over time you definitely improve at typing if you disable it.

I have rather petite hands and fingers, I should note.

Post reply on HN