pejrich

Any model that is going to be meaningfully useful is going to have the potential for harm too. It's not possible to know enough about chemistry to be useful to a chemist, without also being able to figure out how to synthesize drugs. It's not possible to have enough knowledge to be useful to any/all programmers, without being able to figure out how to hack a system, or create a virus. That's just kind of life though.

It's not that simple. The odds are always stacked in favor of the hacker. It's like saying, "why not just make a prison that's impossible to escape from". You can make a prison very, very hard to escape from, but you have to shut off every possible way someone could try to escape, whereas someone trying to escape only has to find one vulnerability, once. It's much harder to plug every possible hole in a complex system than it is to find one point of weakness.

China can produce chips at "7nm", that's more like 8-10 years behind. My M1 macbook pro is 6 years old, and that chip was made using a "5nm" process. The advanced US chips are now on "2nm". Explain to me how China is going to produce a computer that beats US supercomputers using 10 year old technology.

Neither one is illegal though. Scraping the internet is as legal as surfing the internet. And what Anthropic calls "distillation attacks" is really just paying for and using the service Anthropic provides. I would think a judge would have the same view of distillation as they do of scraping, if Anthropic doesn't want a subscriber to have access to the model, they are within their rights to block access, but it's not the user's job to refrain from using the service. If their usage is so different from everyone else's, they should be easy to detect and block. If their usage is so similar to everyone else's that it's difficult to detect, then it shouldn't be of any concern.

Ok, and when there's a power outage at my house as happens quite often in CA? When my internet goes out about once a week and I'm not home to reset it, what then? Just tell the customers "Sorry, I'm running this out of my garage, it's the best thing to do".

I'm not asking them to give me something for nothing. The whole point is to get me hooked and spending so that when those credits run out or a year is up, I'll have to start paying cash. That's like saying giving someone free heroin is just giving them something for nothing. No, the person giving the heroin absolutely has a reason for doing it, and it's not for nothing.

Nope. No, it's not.

Appears so. I was able to start CPU instances.

Seems like a bait and switch when instead of saying "We can give you that SKU in this data center" they say "We can give you that SKU if you use a paid plan".

You must be using some of the higher priced GPUs(which they offered me), but I'm looking for something in the <$1k/month range.

> contributor to Livebook and Nx here

While accurate, it's a bit of an understatement :) Thanks for all your work, Jose.

I'm just finshing up my first macOS app. I went with SwiftUI because I was new and didn't want to learn an "old" tech. The reality is outside of very basic stuff, SwiftUI is not capable of doing a lot of stuff, so I ended up have to learn not only AppKit as well, but how to fuse AppKit into SwiftUI and fuse SwiftUI into AppKit.

My advice is don't be attracted by the shiny toy that is SwiftUI. It's a steaming pile to dung, unless you just want some basic forms/tables/text.

Google is progressively getting worse, month by month. I think Bing will overtake them if they just stand still, but it would be a while. At this rate, I think Google could drop the top spot in 5 years or less.

If Bing is smart I think there's a real chance to dethrone Google in the coming years. In the past 3 years, Google has gone from damn near perfect, to damn near useless. It no longer even listens to what I type in, and very basic queries that I know should have results come back with total meh. I'm not sure what's happening over there at Google. 5 years ago I tried to switch to DDG, but couldn't because the results were too bad. Now, I think DDG is about on par, and that's more from Google degradation, than DDG improvement. It's unfortunate because I'm struggling to find simple things on a daily basis, especially software related things.

Every piece of art is a collage of other pieces of art. Everyone is inspired by other, and copies others, whether they admit it or not. You can't write a song or paint a picture and say that the way you stroke a brush or strum a guitar is not something that was learned by watching others. I think in general too often people are awarded copyright judgements when they shouldn't be, however in this instance I do feel that absolutely nothing of artistic value is being added, and it does seem to be intended to replace the original. If you like a song you heard, and you write something with the same chords, and a similar melody, that's fair game, but if you write something so similar, _and_ market, and sell it as a replacement "don't buy that expensive song, this ones basically the same, and half the price", then I'm quite sure you're in some very murky, and potentially illegal waters.

Well, and here enlies the crux of the issue, can a non-human own or break copy right. You know not too long ago a guy gave a camera to a monkey, the photo that money took became popular and he wanted to capitalize, but the US copyright office clearly stated that they only copyright artistic works created by humans, not those created by non-humans. And it would extend that if non-humans cannot hold copyright, then they cannot break copyright.

Every human artist that ever lived(to my knowledge), heard, or saw someone else create a similar piece of art, from which they were inspired. If I create a song right now, how is that any different than an AI doing the same from being trained on copyrighted music. Certainly my song will be entirely made up of elements i've heard before, however large or small. An ML model is doing the same thing. There is nothing truly original in art. Artists are just filter and amplifiers of what they've heard, seen, and like. Your copyright does not permit you to restrict others from being inspired by your work, or using it for inspiration.

I imagine similar to the use cases for Erlang. Erlang can be a polarizing language, but what it's designed to do, it does exceptionally well, in fact, maybe the best. And that's not concurrency, multicore programming, or distribution. That was all a consequence of the main goal, which was fault tolerance. If you're a Swedish telecom company, you don't want a little bug in the code, that happens during someone's phone call, to knock out a switch, which takes down 50k phone calls, only for some guy to have to trek out into the middle of the forest to restart the computer running everything.

So the model they came up with was high isolation, via processes(no OS processes, but super tiny, like a few hundred bytes processes, all managed by the VM), and distribution(can't have reliability if one node can knock out the system).

The other unique part is the supervision trees that manage those processes. Let's say you have a chat server. You may have a process per device connected to the server, a supervisor for each chat room(group of processes), and a supervisor for all the chat rooms(supervisor of supervisors). Supervisors are themselves processes, but with a specific job.

If your phone's connection/process on the server has some issue, it'll just crash, that's the erlang philosphy "let it crash". You as the user likely can't fix it. The code can't magically fix itself. But presumably through testing we know that initial connections with a clean state are stable, so let it crash, your phone reconnects. Then it crashes again. Reconnects. Crashes again. (This isn't the ideal state obviously, but a worst case for demonstration). Eventually that'll trigger some limit, where the chat group supervisor will notice, somethings wrong here, there's some state that's in a mess, it's time to crash so we can get a fresh try at this. So it'll crash. Usually the first phone reconnection would have been enough to fix it and the group supervisor wouldn't need to restart, but for demonstration purposes we'll say it does, and let's pretend when it comes back up it crashes again. This will propagate up, until enough restarts have fixed the state of the system to the point where it's stable again.

I believe Erlang was the first system to achieve nine 9s of reliability, which is 31.56 milliseconds of downtime a year. Not saying this lib will get you that, but that's the spiel for Erlang.

Well then, let us all break out our extra small violins, and play a somber tune, for the person in their late 30s who still has to work 20 hours a week.

I agree. That’s why my focus is not on being native sounding but conveying meaning. If I say “I want to buy a seal.” The most native sounding translation that says “I want to buy a gasket” is useless if my meaning was “I want to buy a (marine animal).” Obviously that’s a blatant use of an ambiguous word but there’s a surprisingly large number of words we don’t think about as being ambiguous but translators can alter the meaning.

I could be interested. My app is currently translation based(word alignments, sense disambiguating) but had planned to venture into assisted writing for people who need translation assistance, but not 100% help, then giving them a score regarding how much help they needed, or how accurate their portion was. Then using that to help give advice on what they struggle with. Language is hugely complex so I’m currently focusing on ES-EN(or is that spa-eng? Depends on if you’re using nltk or spacy).

I think it could be heaps clearer. To almost everyone outside of a small slice of people in the world, there's a bit difference between "The wallet was hacked and funds were successfully removed from it", and "The wallet was hacked and a key and salt were recovered".

Just because something CAN be done, that doesn't mean it's trivial. Someone's PK "can" be guessed, it's just incredibly unlikely. So for people, like myself, not 100% knowledgable on this type of tech, if someone goes through all the trouble to get the keys, but doesn't complete the final step of actually removing coins, it looks like either for some reason they can't or they are just completely naive and assuming everyone else knows exactly what they've done.

Then perhaps I'm naive about how all this works. But to me there's a difference between accessing keys, and accessing funds. Are these the exact keys, in the exact format, that someone could use to access the coins? For example, if someone were to "hack" my site, and get access to hashed passwords, this would be seen as a breach, and my system would be described as insecure (which is fair), and it could cause issues for people, but getting a hashed password is not the same as getting a plain text password. Are the keys they recovered from the device the equivalent of a plain text password, or a hashed password? Let's say I create a hardware wallet that stores 2 keys (123 and 456), and it uses those two keys to create a private key by doing stuff to it. So 123 and 456 create the pk 132435. Of course this is a terrible system, and not safe at all, but my point is, accessing 123 and 456, is not the same as accessing the actual pk (132435).

If Java is open and free, then why did Android decide to rewrite Java? I notice a lot of companies seem to avoid Java so there must be some sort of a catch.

I agree with McAfee here. Just because they accessed something they weren't supposed to, doesn't mean it was hacked. The system fails when someone gets access to the coins, because that's what the system was designed to protect. I'm not claiming McAfee is correct about anything else, or that the system really is secure, but the point stands that a system designed to secure coins, can only be considered a failure if you manage to extract coins.

Why? I can't imagine a corporation the size of Google would decide to not buy a company, purely based on what language/framework it uses, assuming that buying the company otherwise makes business sense.

I can't think of anything from Rails that I can't do in Phoenix (i've been a Rails dev for 6 years, and working with Elixir/Phoenix for 2). Of course when it comes to the more exotic or rare libraries, they might be easier to find gems for in Rails, than in Phoenix, but in my experience Phoenix has most of it these days. I would say as far as simplicity and dev speed goes, Phoenix and Rails are pretty similar, and Ruby and Elixir are pretty similar. Elixir has a slightly steeper learning curve than Ruby, and Rails has more tuts/docs/etc. than Phoenix, but that's about it.

What exactly makes you say that Rails is a clear win over Phoenix for CRUD stuff?

While it's true that Elixir certainly wasn't built from the ground up, and it has benefited heavily from the work of Ericsson on Erlang, Elixir is more than Erlang with different syntax. A lot of stuff in Elixir was rewritten so that people can read the source in Elixir, and not have to know too much about Erlang.

YES!...YES!...YES!..."MBP's with the touch bar" close window Fuck that.