jdsully

Its an open question if model weights are even copyrightable. So the license may not mean anything.

Phased arrays are their own lense.

Physical hardware products shouldn't lose features after launch. If this was a "mistaken" feature which they suggested it was they should have disabled it on future chips.

"consumers don’t shop based on the endurance metrics even if they should"

Its been well over a decade now and neither I nor anyone I know has ever had an SSD endurance issue. So it seems like the type of problem where you should just go enterprise if you have it.

But surely you see the core LLM innovation is that computers can now TALK to you.

I sign up if I want to do a long road trip and cancel after. Worked great for that.

They hadn't bothered to add ipv6 support to most of their services and the ones that did have it usually were only dual stack - still requiring an ipv4 address.

Wickard V Filburn is simply bad law... The federal government has way too much power and MUCH more than was ever intended. Its not a feature it is indeed a bug. One caused by the Supreme Court.

Excel needed the x87 as well as they cared about maintaining the 80-bit precision in some places to get exactly the same recalc results. So they would have fixed it eventually most likely.

Ah but see the most important piece of information is not what the manufacturer specifies. Most mechanic friends would tell you manufacturers are over-extending the interval to make their cars look good to purchasers and because they only care about getting to the warranty end not total life of the car. While 3k miles old wisdom is out dated, if you do your own oil changes you can see a massive change in what comes out after 5k miles.

By over specifying the question you will miss out on the more important context.

Ask a mechanic friend how often to do an oil change and they will 9 times out of 10 give you an answer without asking what model of car.

You'd still want in transit encryption. There are other methods than centralized trust like fingerprinting to detect forgeries.

At some point it makes sense to just let us use self signed certs. Nobody believes SSL is providing attestation anyways.

The most likely culprit was talking to other nodes via their public IP instead of their local ones. That gets billed as interent traffic (most expensive). The second culprit is your database or other nodes are in different AZs and you get a x-zone bandwidth charge.

Bandwidth inside the same zone is free.

It looks like your writing a letter. Want help?

I wrote Sum Buddy using a variety of AIs. Its a full featured AI spreadsheet. It started in Gemini's web interface and moved over to claude (which was a huge increase in capability). It has a bunch of paying customers now.

https://sumbuddy.net

To be fair they had to lift the plane off the bottom of the Hudson river. It was literally under water. I'm not sure what your expecting in that case.

I was a bit surprised by this take, because I never questioned you would get the luggage back if it survived. Having looked into it more I was pleasantly surprised that even in cases where the airplane was severely damaged luggage was returned.

If you remember the "Miracle on the Hudson", they actually carefully dried everything and couriered it back to the owners. Far beyond what I would expect.

https://abcnews.go.com/Travel/story?id=7629396&page=1

Anecdotally I’m starting to see a more people switch to my spreadsheet app. Not something that should be possible if the MS ecosystem was healthy.

https://www.sumbuddy.net

It was different where you were but the billG and early Ballmer years devs had a lot of power. Like to the point where PMs had training on "soft power" or more bluntly how to get a dev who doesn't want to do your feature to implement how you designed. A good dev certainly wasn't going to get into any trouble refusing a "brain dead" feature in those years.

AWS is degrading on all fronts. If your a large scale business you almost certainly hit stock-outs during peak periods - the elasticity is a lie. If you are a mid scale its high prices and complexity, and finally if your just a beginner good friggin luck getting anything done now that they gutted free support and put near useless quotas on anything usable for modern AI development.

I'm working on Sum Buddy an AI spreadsheet. I didn't like the way Microsoft and Google were integrating their AI by essentially tacking on a chatbox and I wanted to explore more native integrations, like its another part of the tool bar.

https://www.sumbuddy.net/sumbuddy.html

Even if you have that you'll find AWS is "out of stock" and wants you to create reservations that essentially cost the same as just having the machine 24/7.

One word: Ads

There is a truth defence to libel in the USA but there is no good faith defence. Think about it like a traffic accident, you may not have intended to drive into the other car but you still caused damage. Just because you meant well doesn't absolve you from paying for the damages.

The one thing I never understood about these warnings is how they don't run afoul of libel laws. They are directly calling you a scammer and "attacker". The same for Microsoft with their unknown executables.

They used to be more generic saying "We don't know if its safe" but now they are quite assertive at stating you are indeed an attacker.

Apple was failing so their current culture wasn't working and should not be passed on. He wanted a clean break. Plus Apple's past experience was traumatic for him personally. Even if they had been successful without him he'd not want the reminder for that reason alone.

500 usd = 700 cad. So probably more an imprecise currency issue.

Need to test this with my spreadsheet https//www.sumbuddy.net but my users have been getting good results.

A few tips:

- Make tasks short and break them into smaller steps. E.g. don't say "Add a UI button, and a handler that does the thing". But first add the button, confirm its showing as expected then move on to a handler, and so on.

- Do give warnings like "Don't modify unrelated code" but again what the model considers related and what you consider may not be the same (this is much easier if the tasks are small so see point 1).

- If the model keeps making similar mistakes or repeating the same broken thing its because it doesn't know how to solve the problem. These models don't have a way to tell you "I don't know" so they will just keep producing busted code. Give the model additional information to help it like you would a coworker who can't seem to make progress.