No phones can run “LLMs” currently because by definition, large.
Some Android phones however can and does run smaller models locally. Gemini Nano runs on Pixel 8 and can run on Samsung phones.
No phones can run “LLMs” currently because by definition, large.
Some Android phones however can and does run smaller models locally. Gemini Nano runs on Pixel 8 and can run on Samsung phones.
It’s not a LLM, it’s a much smaller model (~3B) which is closer to what Microsoft labels as a SLM (Small Language Models, e.g. MS Phi-3 Mini).
https://machinelearning.apple.com/research/introducing-apple-foundation-models
In both cases I think they would deny it.
I don’t have a solution for the 2nd situation (and I don’t think anyone really does), this is more of a social problem. The point I’m making is that in the first scenario there is a clear pattern of denial for those who need help.
In terms of how you differeniate it, my understanding is if the behavior is impacting your life negatively then you would seek help. But I’m not an expert and that’s not a problem we are solving here.
Up next after Elongate
You can freely think that I am one, and I genuinely think I’m not. If you think that I am one, that’s your problem, not mine.
I mean that’s exactly the issue lol. You might not be an alcoholic, but if you were there’s a solid chance you’d deny that fact.There is a good reason why parts of the 12 step program involve admitting and recognizing that there is a problem, and it’s not limited to alcoholism but substance abuse in general.
A lot of people who genuinely need help refuses to see or admit that they do.
It’s because denial is part of being an alcoholic. The same goes with being called a liar.
make it available to anyone
To do what?
What happened to writing the “core” of an app that doesn’t rely on UI then simply writing the front ends for each platform you want to support?
What do you mean? I can’t speak for Slack but I’m sure some degree of business logic / client side logic separation exists.
By the way, what you just described is the essence of cross-platform development, rather than an argument for building apps natively.
simply writing the front ends for each platform you want to support?
But why would you rewrite the “front-end” for each platform if you have one you could just port over? Who is going to pay for those 2x developers and what would be the ROI on this effort?
That’s just three (if you don’t write for a million desktops on Linux).
Is it really so hard to support just three environments with only the UI being tailored for the OS it’s running on?
In Slack’s case I’d wager the answer to be a resounding YES. I don’t think you fully grasp the full scope Slack’s capabilities, and the amount of work involved to build native clients for not just one or two, but three different platforms - it’s definitely not just the “UI”.
Honestly, it just feels like poor tooling and a poor excuse.
Quite the opposite - frameworks like Electron let’s devs with your skillset build with the stack you already know, and abstracts away quite a bit of the cross-platform complexities, which strangely enough is what you are suggesting but also what you are arguing against
It has all this support for native platforms yet it’s always a clunky memory hog
Maybe so but it has improved a lot over time. The app devs share some responsibility too so it’s not all on Electron.
zero effort to respect the design language of the OS it’s running on.
That’s the Dev’s design choice, not a limitation of Electron.
I’m on macOS, I want the app to be a native macOS app. If I wanted it to look like a webpage, or Windows, or Linux GTK then I’d switch to one of those and expect it to match those paradigms.
I don’t disagree but at the end of the day it doesn’t matter to enough people for it to become an issue. People are used to Slack and the way it works.
Moreover the cost of building the same app 2x or 3x simply doesn’t make business sense.
Flutter came to market much later. It wasn’t even a thing when Slack started building using Electron. I’m sure the same applies to Tauri as well.
For Slack it does. Building an app via Electron means it’s cross-platform by default, so Slack doesn’t need to invest in separate platform teams to solve the same problem (Windows, macOS, Linux).
Electron also has better support for things like native notifications, video and voice calls, offline capabilities, and to other native APIs etc that are either unsupported or spottily supported via the browser.
Calling Slack a webpage is like calling an office building a room.
Slack is just as much a complex app as anything else even if it’s built on web tech and standards.
??
Slack IS an app.
How do you tell if a piece of work contains AI generated content or not?
It’s not hard to generate a piece of AI content, put in some hours to round out AI’s signatures / common mistakes, and pass it off as your own. So in practise it’s still easy to benefit from AI systems by masking generate content as largely your own.
This is such a bad take. By the same logic, the time you spend scrolling through Lemmy would have been better spent serving the homeless in the soup kitchen or sweeping the streets clean. But yet here you are. Why?
Let’s face it, the “IAP money” you referred to has always been there, Sync dev found a way (and has the balls) to monetize his work to a pretty big extent. And only now are you all sour grapes about it. If Lemmy instances are so desperate for money why didn’t they make an equally big effort to monetize? On the flip side, if money isn’t such a big issue why do you care if Sync is monetizing?
sometimes include malware
Source? That’s a pretty serious accusation.
He’s also played Cheney and Gorr.
Medicine can help bring you relief from illness, but it can’t make you healthy if you were never healthy in the first place.
I do think it’s a useful distinction considering open models can be more than 100B+ nowdays and GPT4 is rumored to be 1.7T params. Plus this class of models are far more likely to be on-device.