TL;DR: Apple’s upcoming iOS 18, macOS 15, watchOS 11, and tvOS 18 updates are expected to be “ambitious and compelling,” featuring generative AI improvements, especially in Siri. The revamped Siri may offer enhanced integration with Messages, auto-generated Apple Music playlists, and productivity app integration. Apple is investing $1 billion annually in AI research, aiming to compete with Google and OpenAI. The company is debating whether to deploy the new AI technology via the cloud, on-device, or a combination. (via ChatGPT)
I’ve yet to find a product general AI helped.
I don’t want an AI to make Music playlists. I want simple rules that make sense. Pandora figured that out 23 years ago.
Ai upscaling is a good example. Also chatgpt if you do lots of searches.
The only use I have found for ChatGPT is to generate replies for spam texts.
Why would you want to reply to spam?
Cheap thrills.
It’s just bots adding you to more lists when you reply….
I’ve used it to create plenty of bodgy powershell scripts for home use. It’ll get it 90% right, leaving me to quickly parse and fix it up.
Saves quite a bit of time.
As a C programmer, asking ChatGPT python questions has been super helpful. I can reference some random python library from the corner of the internet and it’ll answer back like it’s an expert in the field. It’s not perfect, but I’m not going to waste my time writing python algorithms if ChatGPT can do it in 2 seconds.
Upscaling is more successful. Using it for search is a very mixed bag, leaning more on the bad side personally.
In 2005 or so, I got a tip about an application called LaunchBar, which would later be copied by Apple to replace the Sherlock search tool, and later by Microsoft in its PowerToys suite. The machine learning LaunchBar used to tailor its responses based on my previous behavior was life-changing. Instead of configuring an application, I just had to use it to change how it behaved.
This is how language models and AI are going to improve your products. Subtly. Behind the scenes. Slightly improving a thousand different use cases, only a fraction of which your regular usage patterns are going to intersect with.
Not sure if you meant general AI or something else, that’s probably going to be a huge step if we ever get there. But the LLMs we currently have are already quite impactful even with their limitations. I don’t think anyone can deny that - and we’re clearly just at the beginning of this.