761–770 of 1001 posts
Earlier quoted context omitted.
What are you talking about? People love Macs for running local LLMs.
For real work tho? My colleagues couldn't get past toy demos. And it ruins battery life. For coding it's on par with GPT3 at best which is amateur tier these days. It's good for text to speech and speech to text but PCs can do that too.
Earlier quoted context omitted.
Apparently the Windows exclusivity period has ended, so Google will support Android and ChromeOS on Qualcomm X2-based devices, https://news.ycombinator.com/item?id=45368167
I mean if the experience is as good as any x86 laptop I'd try it in terms of installing any linux distro I want. No interest in android/chromeos myself.
Earlier quoted context omitted.
Apple has also seemingly stopped caring about the quality and efficiency of their software. You can see this especially in the latest iOS/iPadOS/macOS 26 versions of their operating systems. They need their software leadership to match their hardware leadership, otherwise good hardware with bad software still leads to bad product, which is what we are seeing now.
It is frankly ridiculous how unintuitive it was to add an email account to Mail on iOS. This is possibly the most basic functionality I would expect an email client to have. One would expect that they go to their list of mailboxes and add a new account. No. You exit the mail app -> Go to settings -> apps -> scroll through a massive list (that you usually just use for notification settings btw) to go to mail -> mail a…
The M5 MacBook Pro still gets the Broadcom WiFi chip but the M5 iPad Pros get the N1 and C1X (Sweet). All in all, apple is doing some incredible things with hardware. Software teams at apple really need to get their act together. The M1 itself is so powerful that nobody really needs to upgrade that for most things most people do on their computers. Tahoe however makes my M1 Air feel sluggish doing the exact same task…
Not going to happen, but I can dream.
Does it run Linux?
Earlier quoted context omitted.
Assuming you've read https://andymasley.substack.com/p/a-cheat-sheet-for-conversa... or the longer full essay/related works, could you elaborate on why you don't use Apple Intelligence? I totally understand why someone would refuse to use it due to environmental reasons (amongst others) but I'm curious to hear your opinions on it.
Not sure why would one think that article is something other than distraction attempt. Because emissions are adding up. I'm from country (in Europe) where CO2 emissions per capita [0] are 5.57 while number for USA is 14.3, so reading this sentence in that article: "The average American uses ~50,000 times as much water every day..." surly does not imply that one should use ChatGPT because it is nothing. If "average Am…
Base models only: - M1 | 5 nm | 8 (4P+4E) | GPU 7–8 | 16-core Neural | Memory Bandwidth: 68.25 GB/s | Unified Memory: 16 GB | Geekbench6 ~2346 / 8346 - M2 | 5 nm (G2) | 8 (4P+4E) | GPU 8–10 | 16-core Neural | Memory Bandwidth: 100 GB/s | Unified Memory: 24 GB | Geekbench6 ~2586 / 9672 - M3 | 3 nm (first-gen) | 8 (4P+4E) | GPU 8–10 | 16-core Neural | Memory Bandwidth: 100 GB/s | Unified Memory: 24 GB | Geekbench6 ~296…
Let's see if I can turn this into an ASCII table and have it survive HN's reformatting. +------+------------------+--------------+----------+----------------+-------------------+-------------------+---------------------------+ | Chip | Process | CPU Cores | GPU | Neural Engine | Memory Bandwidth | Unified Memory | Geekbench6 (Single/Multi) | +------+------------------+--------------+----------+----------------+------…
M5 is 4-6x more powerful than M4, which was 5x more powerful than M3, which was 4x more powerful than M2, which was 4x more powerful than M1, which itself was 6x faster than an equivalent Intel processor. Great! Looking at my Macbook though, I can say with utmost certainty that it isn't 4000x faster than the Intel one I had 5 years ago. So, where is the disconnect here? Why is actual user experience not able to keep…
You know, 64% of statistics are made up.
Earlier quoted context omitted.
my tinfoil-hat theory is that on each OS iteration Apple adds a new feature that leverages the latest chips hardware acceleration features and for older chips they do software-only implementations. they ship-of-thesseus the crap out of their OS but replacing with parts that need these new hardware features that run slow on older chips due to software-only implementations. I got the first generation iPad Pro, which is…
At some point you have to use the new features available to you. That's not really tinfoil, just progress, and how all tech works no.
Earlier quoted context omitted.
MLX is a training/research framework, and the work product is usually a CoreML model. A CoreML model will use any and all resources that are available to it, at least if the resource fits for the need. The ANE is for very low power, very specific inference tasks. There is no universe where Apple abandons it, and it's super weird how much anti-ANE rhetoric there is on this site, as if there can only be one tool for an…
> the work product is usually a CoreML model. What work product? Who is running models on Apple hardware in prod?