Number of times AI mentioned: 0
I may be jumping the gun, but this gives the impression that Apple is way behind or doesn’t consider AI a primary business for them, or they have an entirely different approach to AI that is not ready for prime time. My guess is it is the latter.
AI directly no - but LLM’s were certainly mentioned.
Number of times AI mentioned: 0
Just not true, I am afraid. There was quite an explicit emphasis on the ML capabilities of Apple Silicon to do things like use transform models on device in real time and for the M2 Ultra to be developing the models in the first place. Those capabilities already being applied in text correction and dictation and predictive text and lots and lots about ML being used so that your system (whichever one it is) knowing what you are most likely to want to see or use at particular times and in particular places. I’d bet that the suggestions part of the new Journal app (and available to 3rd party developers) is pretty pure application of AI too.
I can understand Apple not wanting to make promises that Siri might not keep, but the obvious place to apply that is making Siri much better and there were strong hints that they were working on this - and a couple of asides that it was about applying Apple Silicon’s power to use AI models of language.
True, there was no promise of “generative” AI for images or words. That’s probably a good thing until someone figures out how to stop generative models creating pure BS.
Cutting edge AI runs on massive GPU servers not personal devices. NVIDIA has done for servers what Apple Silicon did for PCs. Massive computing increase using a fraction of the power.
Indeed! Over lunch today, I commissioned two of my senior leaders to develop a proactive, positive, yet preventative (meaning how do we ameliorate the abuse of AI to preserve academic integrity) assessment and potential development of AI-related courses and teaching and learning structures to support the positive use of AI and deal effectively with the implications (good and bad) of AI use by students and staff.
Apple is doing it exactly right. They’re use machine leaning and not some nebulous “artificial intelligence.”
Unless I missed it, I was surprised there was no mention of CoreML & CreateML in the State of the Union…but they are so capable and underused so far…
I think they do have a different approach.
What is Apple doing with those gazillion neural transistors (or whatever they call it these days) deployed on their Apple Silicon CPUs? As of yesterday, we have these reportedly great realtime personal avatars, we now have transformer-based speech recognition, and I’m sure that GPT-like abilities will arrive at Siri sooner than later, all running on-device. There is AI on almost every product feature one way or the other.
They mentioned “transformers” and “LLM” a couple of times, but that’s a passing mention for what is the rage in technology today. Like with Dropbox on its day, they consider AI as a feature, not a product. It’s the same as with other foundational technologies: almost nobody cares if this or that application is running a SQL database or uses MongoDB or if it’s written in Swift or Javascript, consumers care about features.
There is currently no “AI” on the market!
So Apple could not be behind, because there is nobody at the front…
Although Apple obviously is in part a software company, hardware is by far more key to their business plan.
There is no deficiency in terms of their hardware facilitating AI apps written by others.
“The global Artificial Intelligence (AI) market size was valued at USD 67412.73 million in 2022 and is expected to expand at a CAGR of 39.73 Percentage during the forecast period, reaching USD 501813.57 million by 2028.” - MarketWatch Jun 05, 2023
And who said, that the one who wrote this article has a clue what “AI” really means!?
Seriously? Companies are building AI data centers. Everything is changing.
DGX GH200 Supercomputer, Grace Hopper Superchips in Production
This is not „AI“!!
This is a clever written (complex) software, but has nothing to do with Inteligence!
it doesn’t help that there is no agreed definition of what is and is not AI - it’s a rather broad and rapidly developing field. Apple talked a lot about some aspects of “AI” in the keynote, but didn’t use the phrase once. I think that’s wise - much better to talk about what your OS and systems can do for the user (better text correction, voice recognition, generating avatars in real time etc.) than to jump on an ill-defined hype bandwagon.
AI is a buzzword. Apple calls it “machine learning”, which I think is a more accurate term. They’ve been using that term long before the current hype cycle around LLMs.
Correct.
"At its simplest form, artificial intelligence is a field, which combines computer science and robust datasets, to enable problem-solving. It also encompasses sub-fields of machine learning and deep learning, which are frequently mentioned in conjunction with artificial intelligence. "
@bmosbacker are you following Ethan Mollick at Wharton? He’s writing about AI in education in a useful way.
I have not, but I will certainly check it out, thank you for the suggestion!