Why is Safari not one of the apps? Or is it a Discourse/reddit/etc problem?
Iām not quite sure of the context for the question. Youāre not seeking to proofread text on a web page, are you? Apple Intelligence (given a supported device) will, for example, provide you with a summary of a web page in Safari: see this Apple page.
Stephen
You said:
You cannot do the highlight/click though on text in Safari with the proofread tool. It is all or nothing. So I assume you are saying that Safari is not using the ālatestā text engine on macOS. Is that Appleās fault, or are web designers supposed to include this new text engine? (I obviously do not know how this works.)
In other words, why doesnāt a popular Apple Mac app support one of its key new features?
It literally says ābetaā in image playgrounds and there are lots of similar caveats.
Ha, you are correct! I just went to the official AI site and on the top is an asterisks and the very fine print at the bottom of it says ābeta.ā This of course a footnote on a long screen hyping everything it can do without ever saying beta. Apple marketing should probably learn what beta means.
1 Like
Iām not happy with the Apple-Intelligence-branded releases so far. I think their primary purpose is to placate users and investors until they can release the real Siri replacement and fully use app intents/actions, and get up to speed on deep learning models. Some of the behind-the-scenes work with privacy, cryptographic verification of their ML server units, etc. are cool, though.
1 Like
I think Apple forecasted a tsunami of AI momentum moving away from their products and services so they urgently inserted themselves into the public conversation before anything was ready to deliver. Then they couldnāt deliver on their promises, so theyāre playing catch up with these piecemeal drops. I think all of the features are interesting, some of them are promising, none of them ready for primetime.
1 Like
TLDR; Itās too early to know whether Apple will be successful at implementing AI in the medium and long term, but rejecting what theyāve done so far, and what is currently on offer, is probably foolish.
Apple has been heavily involved and delivering on aspects of what it tends to call āmachine learningā and āneural processingā for a long time. Apple silicon has genuine āneural processorsā in every chip. Thatās all been quietly working away inside their various OSes for years doing things like recognising faces and objects in photos, allowing text dictation and so on and there have been quite sophisticated things going on to make memojis, centre stage in Facetime, adaptive spatial audio and so on. Every photo application has ben using AI features (and the best ones using neural computing) for years now. Thereās been quite a lot of powerful āAIā in Apple systems that ājust worksā rather than being a separate product or identifiable āfeatureā.
They were partly caught out by the mega-hype around large language models, generative image and video systems and, to some extent, by the demand for these. A lot of the hype has come from the particular model of venture capitalism that we are used to these days: very rich private individuals try to guess what will be the wave of the future and invest heavily to try to own it before it actually works.
There have been a few articles recently (e.g. in the Atlantic and Wired) quietly pointing out that NONE of the specific AI companies is remotely near any return on the vast investment that has been poured in, that there is a growing realisation that we might be about to hit the buffers on another round of AI development and hype: weāve had several cycles since the 1950s where extraordinary promise from the early breakthroughs in AI reaches its limit because there is not a straight line between what AI can do in the present and what we might like it to do. Weāre beginning to recognise new sets of inherent problems that have to be solved, that no-one knows how to, like the fabrications that LLMs will produce because you can follow patterns in language and even public information but have no reality check. Or making a voice assistant that has even basic āreal worldā understanding. These are hard problems.
Itās too early to foresee where the current situation will end up. Iād put a small wager on Apple following its traditional line viz. See what emerges thatās likely to be permanent and useful and making a better set of tools and systems than anyone else to exploit them. Looking at the newly revealed and heavily marketed aspects of āApple Intelligenceā as more than a tentative exploration of what might come down the line seems unwise to me. Apple is investing very heavily to ensure that its hardware and operating systems will meet the demands of AI and do it with privacy protected - more RAM, better processors etcā¦ Iād be awfully surprised if that was just speculative or a general development. Tim Cookās recent interviews suggest that they have a much deeper strategy than they are revealing.
3 Likes