No. They handle that. You get so many credits that get used at different rates depending on the model.

You can burn through your months allowance in 10 minutes using Fable then your employer just gets billed the coverage. I think it’s based on the provider’s API pricing so just using Claude or codex directly is probably cheaper.

If you have it do the work entirely and make PRs itself then Copilot is listed as the author regardless of model. E.g. you can assign it an issue and it’ll go and do it itself. I don’t use that though because it’s bad at it.

I tend to just prompt it to write snippets as part as whatever I’m working on to save some typing.

Github copilot is a totally different product from Microsoft copilot. Github copilot is just a harness for any model. I can use openai, anthropic, google, etc. models.

I have at least 1 colleague who does this. Has a Claude subscription despite being provided with github copilot.

New outlook is awful. When I first got forced to switch it didn’t show you tooltip with where a link was actually going before clicking on it - you know, the thing literally every mandatory phishing training has told you to do!

They have actually fixed that now but it was ages without it.

My current pet peeve is it doesn’t tell me about new emails in shared mailboxes. We have a shared support mailbox and I have to remember to go and check it manually. It doesn’t even update the unread email count until you open the inbox.

Unless it has local maps like mine that are now a decade out of date. It’s completely useless now but i refuse to pay several 100 for the update.

I took an existing thing and put it inside another existing thing. That’s a novel idea, right?

We could put card terminals in everything! Put card terminals in the fridge to pay for food?

I can’t think how this is any better than just using the card you already saved on the account. Even from their point of view, this is an extra step in the payment process.

That’s why all these microtransactions are in their own made-up currency.

My view is regulatory capture is as an existential requirement for them. Otherwise at some point we’ll be running open models on our phones and won’t need them.

Not for nvidia though, always going to need chips

Sometimes if I hit the voice typing on my phone, which I’m pretty sure is whisper, and then don’t say anything, it just comes out with nonsense.

Just did this and also got…

Thanks for watching!

My bank has the inverse as well. If i open the app while on a call there’s a huge banner across the top stating they are not calling me.

I remember talking to family in Australia from UK on the phone in 90s and it costing multiple £/min with noticeable delay.

We get $100 a month and most never come close. Mostly use it for typing boiler plate and code reviews. It’s autocomplete to save some typing. It can’t replace your thinking or you’re going to have a bad time.

We make financial software handling millions of dollars a day. Accuracy is more important than speed. You’d think a bank would know that…

BBC posted an article today that’s literally just quoting an OpenAI blog post about how dangerous AI is.

I did a bunch around me with street complete until it started asking the width of the roads. Is that actually useful?

I started filling in all the house names because nothing here is numbered. It hasn’t improved delivery drivers finding us 🤣

I’ve been waiting for all the new software that is going to displace the expensive incumbents.

Pretty much every industry has specialist software that is really expensive so there’s plenty of choices for a competitor to make.

Me too and it’s probably 15 years+ since i played it. It was just subconscious “kotor” appeared in my head and several seconds i caught up and agreed 🤣

We’re good at pattern recognition. That can’t be bad

Back during covid we sent my wife’s grandparents a smart phone so they could video call. They only had a dumb brick phone.

I forget what model but it was one specifically aimed at easy of use for seniors. We had it delivered so I could set it up. Then delivered it to them ready to go.

It was impossible. Even just answering a call being touch and hold the green circle then slide it 2cm in any direction and release took nearly an hour.

My wife does social care now and a client yesterday had managed to enable audio description on his firestick. This is what i was explaining. Hold back and menu on the remote to turn it on/off. He’d probably done it by accident and now his TV is explaining everything on screen after every button press and he has no clue where to even begin to turn it off. It’s utterly incomprehensible. I’m sure they thought it was a good idea to have a way to turn it on that doesn’t need you to see the screen.

Yesterday describing how to do a thing to my wife over the phone “The menu button” “There is no menu button” “It’s a hamburger icon” “What?” “3 lines” “How is that a menu?”

A better example for me is driving, we don’t let children drive, millions of children have access to cars, there’s no age gate on the ignition.

Ultimately its the parents that stop kids driving.

Ban kids from using it. Then treat it like a public health problem and educate everybody on the problems, how to identify misinformation, keep themselves safe etc. And require the companies to provide the tools parents can use to protect their own kids. Parental controls exist but i feel they aren’t widely used because it’s difficult and complicated.

I’d go further and start regulating the use of algorithmic feeds for everybody but that’s probably harder to achieve.