Forgot your password?
typodupeerror

Comment Smaller, on device models are the future (Score 4, Insightful) 26

Most tasks don't really need the big cloud based models. While on device models will lack the depth, breadth, and speed of cloud offerings, they have a number of advantages:

  • Private data can stay private and not get into the cloud models knowledge.
  • For specific types of tasks with well defined needs, the smaller models will be good enough
  • Costs are a lot less and more controllable since you aren't buying tokens at an unknown rate
  • Cloud models can still be accessed if needed via an API while still not revealing private data; for example you could use the cloud to do a lit search and use the data locally
  • They can be hosted on a private server or on the end user's machine
  • Speed isn't necessarily a big issue for many tasks vs costs; and speed can be bought

Of course, if the silicon gets powerful enough to make local models more commonplace, the AI firms may be more reticent to release stripped down versions for free when they discover it's no longer the first hit is free model...

Comment Re:Enshitifcation marches on (Score 1) 93

If the alternatives weren't worse in terms of costs and features I used, I'd have dumped them years ago.

Mint Mobile. Same phone, same cell towers, no contract, way cheaper. Probably cheaper even than your grandfathered plan.

Except it doesn't offer the international data and texting, something I use.

Comment Re:How will the app up-sell? (Score 1) 93

One of the advantages for actual stores is that the reps can attempt to up-sell a customer with a higher cost phone, or a higher cost plan, or additional services ("Would you like fries with that burger?"). Will the app be able to convince the customer they need to send T-Mobile more money?

Ten million nag screens. An app doesn't care if you close a popup over and over.

Comment Re: Bulshit solution looking for a problem (Score 2) 81

There are hundreds of millions in the table.

It's entirely possible for someone to pay someone to start fires, or, have someone reach out with "proof" they started one and ask for payment. Creating a process that encourages either action is only going to make things substantially worse.

Comment Re:Enshitifcation marches on (Score 1) 93

T-Mobile is on one hell of a downward spiral. It's actually impressive.

Exactly. If the alternatives weren't worse in terms of costs and features I used, I'd have dumped them years ago. I had been on an ultra cheap grandfathered plan so no other carrier comes close, even with their price increase.

Comment Re:Chromebooks really don't count (Score 1) 88

I understand your sentiment, but that'd a lot like claiming UNIX has X percent because the Mac is a certified UNIX system. ChromeOS is pretty far from Linux in terms of how it is presented to the user and most of the users have no idea it's Linux:Inside.

Respectfully disagree. The only way to get to the average user (and therefore, to mass adoption) is to present a functional, consistent UI. Once that is achieved, then and only then will the user appreciate the stability, privacy, and flexibility that the underlying Linux OS provides. Given the diversity of distributions and Linux-based products, for communication purposes I would root for something similar to the "Intel inside" logo. The average user didn't need to know anything about processor design and manufacturing, but the logo was, for some time, a very effective marketing strategy.

What about "Linux inside" or something similar to define this category? It would also work for a smart marketing campaign...

Good points. The challenge is, given Linux' fragmented nature, presenting

a functional, consistent UI

Companies selling boxes don't want to have to mess with OS UI design, they want one that user' can use out of the box and runs the programs their customers want on their boxen. Linux' strength is a san open source software people are free to adapt it to their needs, but that is also a weakness when it comes to marketing because it fragments the market and their isn't a consistent desktop GUI or interface standards. A lot of the stuff just isn't polished, even though it works fine, which is a challenge when it comes to mass marketting. People want shiney.

Comment Re:Phone support is a rotten job (Score 3, Insightful) 148

This isn't going to stop at call centres. If not today, then very soon, there will be nothing an experienced legal secretary and AI can't do that a junior lawyer can do...at a fraction of the cost. All that's needed (at least for now) is a mid-level lawyer to scan their output and sign off on it. If I were a couple of years away from a bar exam, I'd be wondering about my career choice.

Ecxept lawyers have been pretty good at ensuring only lawyers can practice law and give legal advice; and while they will no doubt adopt AI to automate tasks low level staff do now, I doubt AI will be a big threat to them like it will for other occupations; rather it will be a tool to be more efficient whithout billing less.

Comment Re:Customer Rights (Score 1) 148

We're going to need a law that customers have the right to talk to a person because the AI isn't going to be reliable.

Most human call centers I have had to contact aren't much better, just some person making a tiny wage reading from a script. Not there fault, it's just companies want the lowest price so that's what they get.

Comment Re:25 x per sec (Score 1) 57

but that's going to have to run on a motor spinning 25fps in the other direction. It could work if frictionless was possible though. I wonder if gyroscopes, low friction and a counter prop like a helicopter would make that feasible.

Yea, or maybe some sort of induction motor to create a counter rotation?

Comment Re:25 x per sec (Score 1) 57

If it's spinning at that rate couldn't you have the camera synchronized to the spins so it's capturing a frame when it's turned to the same position? Kinda like the mechanism that allowed fighter planes to have machine guns mounted in front of the pilot and shooting between the propeller blades. Motion picture cameras are 24 fps.

An axially mounted camera would not ned to spin. Not invisible, but a much smaller thing to spot.

Comment Re:25 x per sec (Score 0) 57

Designing a drone using this technique that could do anything useful would appear to be a considerable challenge. You would have to add the equipment to do whatever you wanted without unbalancing the drone and also make it work despite the spin. Probably not impossible, but would it be worth the effort?

If it can be made large enough to carry explosive charges it would be useful on the battlefield. You might hear it but have a hard time spottig it until it goes boom.

Slashdot Top Deals

Understanding is always the understanding of a smaller problem in relation to a bigger problem. -- P.D. Ouspensky

Working...