The arms race is well underway!
I really like this podcast. Even though I dont always agree with the host, they bring on knowledgeable people.The technical press is so bad, just a bunch of fawning sycophants who regurgitate press releases and hype, so listening to critical voices is crucial to forming a more balanced opinion.
Also, I think generally speaking the wealthy class is "cashing out" so to speak. They're pulling in record profits but laying people off and raising interest rates? Economically that doesn't make sense. There must be sone other reason they're doing all this, and one of the more plausible explanations is the wealthy class is trying to "put down" efforts by the working class to gain power. COVID, the tight labor market, and the successful unionization efforts lately are scaring the shit out of them.
That's my read anyway.
However, you can fake it with the checkbox trick, where you send the client much more data than you reveal at first and use CSS/HTML to incrementally reveal it. You can also do the meta-refresh thing combined with server-side page rendering to pull in new content periodically.
I *think* I set up SMTP but I'll have to double check that.
I'm not sure I'm comfortable with automatically nuking accounts. I think I'd rather look over the list myself and have the opportunity to modify which accounts are deleted. It seems like good practice to me to not automate administrator decisions that affect people. Yes, the vast majority of these accounts are probably not backed by real people who want to participate in a non-spammy way, but I don't like the notion of accidentally catching a person up in an algorithmic purging.
I was thinking that I rarely use the actual telephone network anymore. It's almost all digital/internet.
I do love SQL as a language concept though. For a grad school project the group I was in wrote an interpreter for a SQL-like language that queried email, and it's so elegant.
The notion as I understand it is that they tell companies, on your behalf, to stop collecting data about you and to delete what data they have about you. Consumer Reports doesn't know this information; they act as an intermediary. They're and old and trusted non-profit in the US so it makes sense: consumers can trust them to act on their behalf, and companies tend to not want to piss off Consumer Reports.
The notion as I understand it is that they tell companies, on your behalf, to stop collecting data about you and to delete what data they have about you. Consumer Reports doesn't know this information; they act as an intermediary. They're and old and trusted non-profit in the US so it makes sense: consumers can trust them to act on their behalf, and companies tend to not want to piss off Consumer Reports.''
Edit: oops I got sidetracked and didn't answer your question lol. Truth is, I don't know what they need to know from you. I imagine if you have IOS and install the app you'd figure that out pretty quickly. I signed up to get an email when their Android version comes out so I can try it.
Interesting on a lot of levels. Consumer Reports is well respected and has been around since at least when I was a kid. If they're actively calling out large companies for stealing and selling people's data the tide has surely turned on this practice.
Wowowowowow, an *encryption key* for *user data* was stolen, too. What the hell kind of clown show is LastPass running over there? If you use that thing get out now!
So it would seem that having a way of transferring control of a project from absent maintainers to aspiring active maintainers would be very nice to have. You don't want that control to pass into the hands of some arbitrary person. But, let's say there are people who are not currently maintainers issuing pull requests; and there are users submitting issues. You could imagine a mechanism whereby the pull request issuers can be "promoted" by community consent, where the community is the people submitting issues and other pull requests, plus anyone else who wants to be involved perhaps.
There are obviously dangers, but I don't think they are meaningfully different from the dangers that already exist in the
github model of software development. Maybe I've overlooked something important though.
github. It really is an annoying problem if you depend on a project where the main maintainers go absent without passing the project on to someone else. The project becomes trapped and dead. Usually (and rightfully), only the maintainers can push releases that can be used by a wider community. But that means if you're depending on a ruby gem or an npm package of a java jar or....then the release package is no longer updated once the maintainers go absent. People can submit pull requests, but with no maintainers to accept them, the source code goes stale too. Though you can grab the pull release(s), the merge process often requires project-specific knowledge that has gone absent with the maintainers.
github. It really is an annoying problem if you depend on a project where the main maintainers go absent without passing the project on to someone else. The project becomes trapped and dead. Usually (and rightfully), only the maintainers can push releases that can be used by a wider community. But that means if you're depending on a ruby gem or an npm package or a java jar or any other build artifact on an official channel, you're out luck because the release artifacts are no longer updated once the maintainers go absent. People can submit pull requests, but with no maintainers to accept them, the source code goes stale too. Though you can grab the pull release(s), the merge process often requires project-specific knowledge that has gone absent with the maintainers.
Currently
Not
Exactly
Truthful
> First, the site was caught quietly publishing the machine learning-generated stories in the first place. Then the AI-generated content was found to be riddled with factual errors. Now, CNET's AI also appears to have been a serial plagiarist β of actual humans' work
Is it just me or is this stuff getting worse at an accelerating pace?
I think UnifiedPush is coming along nicely, but I wonder why they don't talk about IOS. The Conversations XMPP app for Android can now function as a UnifiedPush distributor, which is very cool--that means that any app that supports UnifiedPush can send notifications to you via XMPP instead of using Google's proprietary spyware crap.
systemd service file, @eldersnake
yarnd.service unit on the yarn git awhile back if you use
systemd. It could use some work but a variation of that one is controlling my own pod well enough.
1. run
sudo tune2fs -l BLOCK_DEVICE | grep 'Filesystem created:' on a BLOCK_DEVICE whose filesystem was created at 1st machine use2. run
smartctl -a BLOCK_DEVICE | grep Power_On_Hours to check the total power-on hours of some BLOCK_DEVICE that's been up since the machine's 1st useObviously these both depend on having a block device (disk drive usu) whose life span is close to the machine's total uptime. There are utilities like
tuptime in Linux, which I think are also compileable on OpenBSD, that you can install when you first start up a machine to keep this cumulative uptime but that doesn't help after the fact unless you solve time travel!
I hear what you're saying about scarcity. You don't know me, nor my background, so please be careful with your assumptions on that point. I did grow up in the US, and the US is extremely privileged in many ways, but those privileges are not universally distributed. I'd also point out that I am coming from the perspective of a credentialed computer scientist who feels a responsibility to his field, and I am operating in the context of using words and rhetoric, a tool available to everyone. I don't have degrees from fancy schools like Harvard or MIT that give my words extra weight; if I did I wouldn't be writing this on yarn.social, I'd be writing this in the New York Times.
You seem to be ignoring that not so long ago, nobody knew the word "ChatGPT". Now they do. Somebody changed the world to introduce that word into more people's vocabulary. I wish to change the world to change how people conceive of what that world means. I'm not going to stop trying to do that just because there are 8 bullion people on Earth.
But I don't know what, if anything, can be done about that. Many of my computer scientist colleagues contribute to this! Generally, we're not organized to counter the bizarre claims that are pushed into the media by SV snake oil salesmen.
But I don't know what, if anything can be done about that. Many of my computer scientists colleagues contribute to this! Many don't, but the SV-aligned voices seem to win out every time.
But I don't know what, if anything, can be done about that. Many of my computer scientists colleagues contribute to this! Many don't, but the SV-aligned voices seem to win out every time.
The state of the world is what we make it. As computer scientists, we have a responsibility to hold the line against this maniacal hype. It's insane, in an almost literal sense, for us to follow the herd on this particular issue. We know better. We know all about just-so stories. We know about mechanical Turks. We've been through half a dozen or more AI hype cycles where the latest thing, whether it be expert systems or case-based reasoning or cybernetics or the subsumption architecture or neural networks or Bayesian inference or or or...., was going to replace the human mind. None of them have become that, to a one, and ChatGPT won't either.
1. ChatGPT does not have a conception of what is going on in the world. It is a word-emitter that tricks human minds into thinking it does. In other words, it's a kind of complex automaton, a marionette. The fact that the action of it is complex enough to fool us into thinking it "knows" something does not mean it does
2. ChatGPT is as likely to emit false information as true information (perhaps more so; has this been assessed?)
3. ChatGPT does not have deductive or inductive logical reasoning capabilities; nor does it have any "drive" to follow these principles
4. Human papers are for human writers to communicate to human readers. It seems to me that the only argument in favor of including ChatGPT in this process is a misguided drive to speed up the process even more than publish-or-perish has. In fact it should be *slowed down* and made *more careful*.
5. The present interest in ChatGPT is almost entirely driven by investor-fueled hype. It's where investors are running after the collapse of cryptocurrency/web3. There is a nice interview with Timnit Gebru on the Tech Won't Save Us podcast, titled "Donβt Fall for the AI Hype" that goes into this if you're curious. As computer scientists, we should not be chasing trends like this.
1. ChatGPT does not have a conception of what is going on in the world. It is a word-emitter that tricks human minds into thinking it does. In other words, it's a kind of complex automaton, a marionette. The fact that the action of it is complex enough to fool us into thinking it "knows" something does not mean it does
2. ChatGPT is as likely to emit false information as true information (perhaps more so; has this been assessed?)
3. ChatGPT does not have deductive or inductive logical reasoning capabilities; nor does it have any "drive" to follow these principles
4. Human papers are for human writers to communicate to human readers. It seems to me that the only argument in favor of including ChatGPT in this process is a misguided drive to speed up the process even more than publish-or-perish has. In fact it should be *slowed down* and made *more careful*.
5. The present interest in ChatGPT is almost entirely driven by investor-fueled hype. It's where investors are running after the collapse of crytocurrency/web3. There is a nice interview with Timnit Gebru on the Tech Won't Save Us podcast, titled "Donβt Fall for the AI Hype" that goes into this if you're curious. As computer scientists, we should not be chasing trends like this.
salty uses saltpack. I guess thsrs why it's called salty huh? π€¦ββοΈ
yarn release soon so all this good stuff can be available in a versioned version. π
I've been playing around with this version of Rymdport, which is a cross-platform GUI for
magic-wormhole, and it seems pretty nice. I've been able to send files between several different Linux computers as well as between a computer and my phone with it.One really nice feature is the direct support for sending text snippets. This is a good way to "securely" send something like a password from one computer to another. I say "securely" because you still have to communicate the wormhole code to the recipient, which of course could be intercepted and abused. The upshot though is that (a) you will know if someone connected to and downloaded whatever it was you sent, which allows for some amount of tamper detection; (b) the connection, once established, transmits data and closes without you having to do anything special, so there's no chance that data is accidentally left available to the world.
The
wormhole cli can receive text and sends it to stdout, so you'd probably want to do something more sensible with that if you send a password (like pipe it into a password manager or some other downstream receiver).
yarn.This talk, which is about
wormhole-william generally, has a segment about why he chose to use Gio UI to build the smartphone versions of the app. The link should take you to that spot but if not it starts at the 10:22 minute mark.
yarn.This talk, which is about
wormhole-william generally, has a segment about why he chose to use Gio UI to build the smartphone versions of the app. The link should take you to that spot but if not it starts at the 10:22 minute mark. "I was able to go from no experience to yes I can turn this into a working app in a very short period of time".
yarn.
- Box: "Package box authenticates and encrypts small messages using public-key cryptography"
- Secretbox: "Package secretbox encrypts and authenticates small messages"
- Saltpack: "Need to encode, transmit, or store encrypted or signed data? saltpack is a streamlined, modern solution, designed with simplicity in mind. It is easy to implement & integrate"
- Magic wormhole: "This package provides a library and a command-line tool named wormhole, which makes it possible to get arbitrary-sized files and directories (or short pieces of text) from one computer to another".
The first two are Go libraries so may be of special interest. Saltpack was put forward by keybase, but it's also a format that can be implemented in any language. Magic wormhole is implemented in either
python or rust; the protocol is described here.