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

Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

gizmoweek.com

Post reply on HN
Print view View original

191–196 of 196 posts

Re: Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

#191

Post by subscribed »

cubefox wrote: post #83
subscribed wrote: post #65

Earlier quoted context omitted.

These detectors are a scam falsely flagging non-native English speakers: https://plagiarismcheckerai.app/ai-detector-false-positives-... At this point relying on their judgement is beyond folly.

It's both ironic an confusing that this website itself promotes an AI detector.

Yeah, I admit I lazily chose one of the first results reporting on this study instead on the best one, so the irony is not lost on me.

Sorry for making you snort and shake your head in amusement :D

Re: Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

#192

Post by Forgeties79 »

Gareth321 wrote: post #182
mrkpdl wrote: post #44

Earlier quoted context omitted.

But… why would I put the effort into getting an llm to make me an app when a there’s an existing app that I don’t have to maintain? I don’t want to have to make every app I use?

It depends on the value of your time relative to the price of the application. Many of us here are well paid and time poor. It makes sense to pay for well built apps. For the world, we are truly the top 1%. 99% of people would be happy to spend a few hours building an app they need to occasionally maintain. Especially if the app is subscription based. We should also remember that the effort of building and maintainin…

> 99% of people would be happy to spend a few hours building an app they need to occasionally maintain. Especially if the app is subscription based.

With people slowly abandoning dedicated computers they fully control (if we can even call windows/macOS that) and going towards mobile/tablet computing more every year, which is a far more locked down device run by companies that are becoming increasingly hostile to “side-loading,” I just don’t see how this can become reality.

Re: Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

#193

Post by Gormo »

yangm97 wrote: post #101
figmert wrote: post #47

Earlier quoted context omitted.

> :v I guess I found the millennial. I haven't seen that in so long!

Analog emojis FTW

Neither analog nor emojis. An analog emoji would just be a picture printed on paper.

Re: Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

#194

Post by fsiefken »

dpacmittal wrote: post #15
fsiefken wrote: post #14

Earlier quoted context omitted.

Qwen3.5-9b and Qwen3.5-27b are pretty coherent on my 24G android phone

Which android phone has 24G?

For example: ASUS ROG Phone 9 Pro, Nubia RedMagic 10 Pro, OnePlus 13, Realme GT 5

Re: Google Gemma 4 Runs Natively on iPhone with Full Offline AI Inference

#195

Post by krackers »

Yukonv wrote: post #143
jonathaneunice wrote: post #142

Earlier quoted context omitted.

Are the Apple neural engines even a practical target of LLMs? Maybe not strictly impossible, but ANE was designed with an earlier, pre-LLM style of ML. Running LLMs on ANE (e.g. via Core ML) possible in theory, but the substantial model conversion and custom hardware tuning required makes for a high hurdle IRL. The LLM ecosystem standardized around CPU/GPU execution, and to date at least seems unwilling to devote res…

It is possible but requires a very specific model design to utilize. As this reverse engineering effort has shown [0] "The ANE is not a GPU. It’s not a CPU. It’s a graph execution engine." To build one requires using a specific pipeline specifically for CoreML [1]. [0] https://maderix.substack.com/p/inside-the-m4-apple-neural-en... [1] https://developer.apple.com/documentation/coreml

More info on specific design choices needed to run models here [1]. I mean it is possible given that apple themselves did it in [2], but it's also not as general purpose or flexible as a GPU.

[1] https://news.ycombinator.com/item?id=43881692 [2] https://machinelearning.apple.com/research/neural-engine-tra...

Post reply on HN