I watched a neat if somewhat synthetic ML demo based on this paper - arxiv.org/pdf/2005.11401.pdf - and on the one hand, the technique is interesting because it lets you isolate cited sources for the generated text fairly easily. But on the other hand... I remember when people thought using Wikipedia for research was cheating. And it still - sort of - is, in the sense that it's not your primary research, it's other people's distillates of other people's research. And that's what this demo is doing.
And to take that logic a step further: the reason Wikipedia was considered cheating or in some way adjacent to plagiarism wasn't just its presumed unreliability - some people just need to see gatekeepers, I get it even if I don't agree - but the fact that all this supplementary labor went not just uncited or uncredited but unacknowledged.
That is to say, laborwashing.
Yes, it's novel that ML demos can distill coherent sentences out of the products of a tectonic amount of human effort. But.
So far I haven't seen one of these tools that isn't, in one way or another, glossing over the step where a lot of un- or under-paid humans are doing all the heavy lifting that makes any of this work at all. Do you presume to criticize the great and powerful Oz^H AI model? Pay no attention to the hundreds or thousand of people behind the curtain.
I'm naive enough to believe humane computing is possible, that computational literacy matters, that access to computation should be a human right, but for any of those wild-eyed techno-hippie ideals to matter, computation can't have exploitation at its core. And all the talk about "AI safety" and "human-in-the-loop decision making" is a distraction from the fact that it does, hiding questions about where humans already are in these systems with chin-stroking questions about where they should be.
To some extent, I think our language is failing us here, in the way it always fails us in the presence of real novelty; we seek out analogy to tether the novelty to our existing frames, and the shortcomings of those analogies are where the charlatans of the world do their best work. "We are encoding knowledge" no sir you're encoding text. The knowledge is represented by the text but it is not knowledge. "Signs of AGI" sir you put googly-eyes on a very complicated spreadsheet, we are not fooled.
This is, I think, where my bottomless contempt for the threat-of-AGI, Roko's Basilisk crowd comes from, the callous willingness to overlook the real cost of systematically undervaluing humans in their desire to either saved or destroyed by a God they're trying to bring about out of the fear it would be unhappy if they didn't.
It's just so dumb. It is incredibly dumb, but it's a brand of dumb that comes at huge human cost, built on an abstraction layer that lets believers ignore that cost.
Anyway, the number of self-educated people out there who believe that if you put a dictionary in a blender there's a chance the blender will get extremely angry you didn't do it sooner and seek revenge, so logically we need to put all the dictionaries we can find into the biggest blenders we can build as soon as possible all the while calling themselves 'atheists' is prima facie evidence that funding robust liberal arts programs is a national defence issue.
Berlin, November 23, 2023 – LibreOffice 7.6.3 Community, the third minor release of the 7.6 family of the volunteer-supported free office suite for Windows (Intel/AMD and ARM processors), macOS (Apple and Intel processors) and Linux is now available …
The #LibreOffice Viewer for Android is back on the Google Play store, with a bunch of improvements! It has experimental editing support too – but to make that ready for everyone, you can give our community a hand to finish it: play.google.com/store/apps/det…
I downloaded the Android apk and tried it. As the fallback font and the default font of the app is Noto Naskh Arabic, which is right-to-left writing language, Myanmar Unicode Rendering went wrong. Myanmar Unicode is a right-to-left writing abugida. Not only this app but also the latest version of Collabora Office for Android has this bug or problem. Please remember that on Android, we don't have Control+A to select all and change to Noto Serif Myanmar easily. Please reuse Liberation Serif as the default and fallback font. I think it's pharmacokinetic side-effect of oil-producing Arabic-cetered developing to counter BRICs.
@conjurenation We're volunteers working hard to give you (and the world) a free office suite. You didn't provide any details about what you think is "broken", but if you're not willing to help the volunteers who work so hard to give you free things, I guess there's not much we can do... 😕
@conjurenation Again, we are volunteers working very hard, with very limited resources. That's not a "lame excuse", and is insulting to the people who work really hard to give you a free office suite. Have you considered helping us?
Semi-automate and perform the update to LibreOffice Viewer 7.6.3. Further simplifying/automating the update in the future might be an option, but depends on upstream changes as...
Miluju, jak tělo umí vymyslet krásný fuck-up na moji hlavu. Byl jsem víc jak týden nonstop v práci, tak jsem se dost vykašlal na firmu. A čím víc se blížil dnešek, jsem to s klidnou hlavou odkládal "do čtvrtka to vydrží, máš volno, dozeneš to". Moje tělo dneska:"39.2, rýma, kašel, hlava jak střep"
Zahajují frontální útok. Dvě mandarinky, grapefruit, litrový hrnec čaje s citronem, medem a slivovicí, jdu si pustit Netflix, než budu zase zdravej. Tak
Nemocnice na severu Číny praskají ve švech. Oblastí se mezi dětmi šíří respirační onemocnění není jasné zda se jedná o novou vlnu covidu, jiné virové onemocnění, či zápaly plic.
Jakýkoli autoritářský nebo totalitní nebo despotický režim je nebezpečný celému světu. Tipuji, že jde o nafouknutý prd, ale po vcelku nedávných zkušenostech už se nad tím mávnout rukou nedá. Už i rozumného člověka pak napadají všelijaké konspirace.
The #KOSA Bill threatens to censor online content under the guise of trying to "protect the children". 🤔
The bill's vague language opens the door to dangerous legal interpretations that will do more to push censorship on the American public than protect the nation's youth. 🚫
Hello, I have a feature request, is it possible that you integrate the setting that the bottom or the default top bar disappears when scrolling? (Look at the Image)
Blockqoutes and code blocks in posts now look nicer.
The old behavior of the notification tab (pre Tusky 22.0) has been restored.
Role badges are now shown on profiles (Mastodon 4.2 feature).
The v...
Testing now, looks nice. Good thing the text size can be adjusted. Absolute timestamps are nice. Would be nice if the timeline could be reversed when reading oldest first. The Finnish translation needs some work, especially the publish button which says Tuutaa! which doesn't mean anything. (Tuuttaa! could possibly mean Toot!)
Denní show My jsme Západ přináší vše důležité, co se ve světě politiky a financí děje.Zapomeňte na banky:https://www.anycoin.cz/?ref=bratricekLightning BTC d...
#dobréRáno no už som hore a proti svojej vôli.. spolubývajúci vydával ráno nejaké divné zvuky, som vzťahovačný ale toto bolo divné dosť.. jo a dávam do kopy čo si oblečiem
edit: ale by som ešte spal teda edit1: že som odolal večer nátlakom pitia borovičky pokladám za "feature" edit2: -1
@archos Děkujeme, i když si dovolím nesouhlasit. Podle mě jsou krásný všechny český mastodonmňačky. 😍 Je to dneska už masově chované plemeno mainská mývalí.
I got myself a little N900 a couple of weeks ago. The USB port was busted (completely off lol), but I got some new batteries and an external charger. And now it lives! It's running maemo 5 from 2011 so a lot of stuff just doesn't work. Which is fine. I'm going to replace the usb port and then try maemo leste on it. The fact that a modern OS is available for such an ancient device is so great! #nokia #maemo #linux #linuxphones
Apropos people shot in Panama recently because protests for a mining contract for a Canadian company, and environmental (and political) problems in Costa Rica because another Canadian mining company.
Another classic, incisive post from @mmasnick Re: Elon’s supposed thermonuclear lawsuit against Media Matters:
“So… within the span of about 2 to 3 pages we are told that Elon Musk and exTwitter are passionate supporters of free speech that allow “all viewpoints” to be shared and that Musk is filing this lawsuit to force Media Matters to take down speech that he admits is absolutely true, but where he doesn’t like how they portrayed things.”
So, yesterday we covered Elon Musk’s ridiculous censorial threat to sue Media Matters “the split second the court opens on Monday.” Of course, as we noted, you can file lawsuits 24/7. And yet, as t…
Did you know @kde had an #accessibility inspector? Currently it's hidden in the build folder of libqaccessibilityclient but not for long. We started the process of making it a standalone application and a first release is planned in 2-3 weeks!
@Tutorial Strange is that your account is reported as DFRN contact and not ActivityPub . But that is unrelated to this issue. Just what I have observed.
thx for that observation @Roland Häder, there is definitely something wrong with the forum pages over here, at least with those created by the related account of @Tutorial .
Still trying to define what's wrong for the consequent error or bug report.
This might have something to do with a setting by the previous hosting in local.config.php, there is a post in the helpers forum about that wrong setting. Maybe there is still a change necessary from 'localhost' to '127.0.0.1', I didn't find the time yet to address and test that.
Even tho it is unrelated that the tutorial account and also this one is reported as DFRN contact and not ActivityPub, that might be some hint to something, there are some things that don't work as expected or as they should so this is a hint to have at least in mind.
I just posted this follow up help request with "into deep" information about the CHMOD settings of the smarty3 folder: tupambae.org/display/0ac89072-…
I had a look at the smarty CHMOD settings and found the following:
drwxrwxr-x 3 www-data www-data 4096 Nov 8 20:23 smarty3 If I'm not wrong that's CMOD 775 (rwx|rwx|r-x) (?)
In the installation process the commands to create the smarty folders were: www-data@VPShosting:~/html$ mkdir -p view/smarty3 www-data@VPShosting:~/html$ chmod 775 view/smarty3 see: squeet.me/display/962c3e10-156… tupambae.org/display/0ac89072-… -- I looked into the subfolders and found:
rootname@VPShosting:/var/www/html/view/smarty3# ls -l drwxr-xr-x 222 www-data www-data 4096 Nov 25 17:20 compiled => CMOD 755 (rwx|r-x|r-x) (?) -- The folder "compiled" has a long list of sub-folders apparently each having 2 more steps of sub-folders.
rootname@VPShosting:/var/www/html/view/smarty3/compiled# ls -l total 880
I found two types of folders, some few created on different dates strangely belonging to the user root instead of www-data, here two examples and how those two types of subfolders look like. I guess the folders owned by root are wrong? -------------------- drwxr-xr-x 3 root root 4096 Nov 12 04:35 00 - rootname@VPShosting:/var/www/html/view/smarty3/compiled/00# ls -l drwxr-xr-x 3 root root 4096 Nov 12 04:35 d4 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/00/d4# ls -l drwxr-xr-x 2 root root 4096 Nov 12 04:35 ec => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/00/d4/ec# ls -l -rw-r--r-- 1 root root 675 Nov 12 04:35 00d4eca105abd94437094f3d4409477acb55526a_2.string.php => CMOD 644 (rw-|r--|r--) (?) -------------------- drwxr-xr-x 3 www-data www-data 4096 Nov 22 20:25 01 - rootname@VPShosting:/var/www/html/view/smarty3/compiled/01# ls -l drwxr-xr-x 3 www-data www-data 4096 Nov 22 20:25 97 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/01/97# ls -l drwxr-xr-x 2 www-data www-data 4096 Nov 22 20:25 f2 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/01/97/f2# ls -l -rw-r--r-- 1 www-data www-data 6140 Nov 22 20:25 0197f2d4b23957a898d38870d6c6a3775da487ff_2.file.group_side.tpl.php => CMOD 644 (rw-|r--|r--) (?)
I had a look at the smarty CHMOD settings and found the following:
drwxrwxr-x 3 www-data www-data 4096 Nov 8 20:23 smarty3 If I'm not wrong that's CMOD 775 (rwx|rwx|r-x) (?)
In the installation process the commands to create the smarty folders were: www-data@VPShosting:~/html$ mkdir -p view/smarty3 www-data@VPShosting:~/html$ chmod 775 view/smarty3 see: squeet.me/display/962c3e10-156… tupambae.org/display/0ac89072-… -- I looked into the subfolders and found:
rootname@VPShosting:/var/www/html/view/smarty3# ls -l drwxr-xr-x 222 www-data www-data 4096 Nov 25 17:20 compiled => CMOD 755 (rwx|r-x|r-x) (?) -- The folder "compiled" has a long list of sub-folders apparently each having 2 more steps of sub-folders.
rootname@VPShosting:/var/www/html/view/smarty3/compiled# ls -l total 880
I found two types of folders, some few created on different dates strangely belonging to the user root instead of www-data, here two examples and how those two types of subfolders look like. I guess the folders owned by root are wrong? -------------------- drwxr-xr-x 3 root root 4096 Nov 12 04:35 00 - rootname@VPShosting:/var/www/html/view/smarty3/compiled/00# ls -l drwxr-xr-x 3 root root 4096 Nov 12 04:35 d4 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/00/d4# ls -l drwxr-xr-x 2 root root 4096 Nov 12 04:35 ec => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/00/d4/ec# ls -l -rw-r--r-- 1 root root 675 Nov 12 04:35 00d4eca105abd94437094f3d4409477acb55526a_2.string.php => CMOD 644 (rw-|r--|r--) (?) -------------------- drwxr-xr-x 3 www-data www-data 4096 Nov 22 20:25 01 - rootname@VPShosting:/var/www/html/view/smarty3/compiled/01# ls -l drwxr-xr-x 3 www-data www-data 4096 Nov 22 20:25 97 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/01/97# ls -l drwxr-xr-x 2 www-data www-data 4096 Nov 22 20:25 f2 => CMOD 755 (rwx|r-x|r-x) (?)
rootname@VPShosting:/var/www/html/view/smarty3/compiled/01/97/f2# ls -l -rw-r--r-- 1 www-data www-data 6140 Nov 22 20:25 0197f2d4b23957a898d38870d6c6a3775da487ff_2.file.group_side.tpl.php => CMOD 644 (rw-|r--|r--) (?)
¿Hi there @Friendica Developers, can you please take the theme VIER immediately out of the branches stable and RC!
It doesn't even display the [abstract] bbCode properly and right now literally I have no clue how many hours of my life I spend the last month documenting bugs that might be just because that theme doesn't work.
My mind is raging and I just jumped out of the window, lucky me that I live on ground level so every thing is fine as of now, but please, this is disheartening! thx
true @xy.. , but to be sincere, as of now I'm still trying to sort out what is and what is not with respect to bugReports "and all the rest": tupambae.org/display/0ac89072-…
Once I have my list and reports set and done I guess I will have no other choice than to visit microsofts github to work for their AI, post all the reports at once and run away as always, scrubbing my browser and everything that is not "Niet und Nagelfest".
Perhaps, actually I hope, I will use git.friendi.ca first to post those reports, just because that's how it should be and actually, as of now this community is quite small, I truly consider we should have a bugResearch forum, pin there a bugResearch layout like this one: tupambae.org/display/0ac89072-… and do the basic research and communication here in our "world". Once an issue, or feature request is pinned down, defined and summed up in such a report, it's just a copy/paste for anyone, or even someone in charge of that specific forum, to post it to gitHub and that's it.
There was a moment in time on diaspora where the arousal inside the community with respect to discussion and in particular feature requests was such that it was pure noise, in that moment github or any other instance of more internal discussions for those who know and actually do the stuff was like a protection layer for them, for coders and developers. Actually recently the same happend on mastodon. For now I can't see that noise problem here. I think that there are lot's of people who might help out in certain things, like testing and checking, moderating and or mantaining a forum page, but wont sign up for another site and another site and another site. A sign up that btw means, getting used to another layout, environment and so on.
In other words, I truly believe that this "sign up on github" isn't usefull in most cases. In this particular request much less. It's plain simple, if we can't maintain two themes working correctly, pass one to exclusively dev, make a fork for it or create an addon (best choice) for it. Friendica started with like 5 themes, the one mike started with and was used to for his coding and eyes, plus cuatro, vier, frio ..
If we can't even do it for two, so let it be one and the rest as addons. I don't like the layout of FRIO, I like the VIER layout, but believe me, I just got it, if you actually want to work, create content help with bug search or what ever, just use the theme the main coder uses or die in the intent.
And for the github thing, I was already angry before chatGTP about another centralized mega player like github, even before MS bought it. And when I saw git.friendi.ca for the first time I immediately created an account there. We are creating our fediVerse because of the very reasons of the totalitarian IT system they are building, holding all the keys in their hands. And now, every time someone starts to work with us testing and using the software we are sending them right into the arms of those "A****"?
Sorry but actually that's wrong and we are not even evolving the potential of our own place and software to begin with as I already pointed out.
So thx for the hint, I know gitHub exists, I forked friendica in 2011 over there to the account I have there.
Your server can have it anyway, Nanook that's how github and free software works.
FRIO looks like designed by someone who spent to much time on twitter. I don't like how it looks and I don't like how editing and posting looks/feels compared to VIER but that's not the point and to begin with what I became accustomed to and there for don't want to change in the first place.
This request is about the fact that apparently it takes resources the main coder community doesn't have to make more than one theme work and integrate into the main branch.
VIER (or FRIO), and to that matter any other theme could/should be an addon. Actually if it was it would incentivize others to code optional themes too or modify VIER (or FRIO).
To offer a theme in STABLE or RC that doesn't work/integrate with the basics of friendica like bbCode is bad for friendica, for the users and for the community.
Apparently we are not talking about some looks or outfit here but different coding from the core:
> "Frio" Responsive theme based on a modern HTML/CSS/JavaScript framework. V.1.0
If I expect anything from friendica in the first place it is that it is fast and simple. From that point of view I actually even would want it to be just simple text and bbCode or markdown addons to be activated by admin and user.
Having been raised in terms of websites in the phpBB board world, I'd love to be able and incentivized to tweak friendica with addons to my needs. Manipulate the bbCode options below the text space, for example ad a button for (code), or [p] or [hr]. Or shut down the bbCode proposition that pops up every time I start writing a bracket because that's a "pain in the a**" at least here in VIER.
friendica is not microblogging, all the contrary, so I don't know why the basic theme doesn't just give me as much screen width as possible to write and have basically a letter kinda layout.
@utopiArte The problem is that if they remove it, then the base code will rapidly become incompatible and it will no longer work, that is why I say my site will become unmaintained.
> then the base code will rapidly become incompatible
My assumption is that it is already.
In general terms speaking, if your site has a stable version installation and there are no important security fixes, or changes in the activitypub protocol itself, it should work as is without updating it.
Nanook, if you have a look at the view folder of friendica you'll find that the old themes, even Mikes duopuntocero are still there, so even if VIER falls out of the admin panel it most likely still would be available for profiles that are set to that theme from before.
As you can see VIER isn't already up to date like my recent post about the bbCode [abstract] pointed out: tupambae.org/display/0ac89072-… and there are other things that don't really work when I try to answer in other posts for example. So what you are fearing is what I'm actually pointing out is happening already, two themes are to much work for the dev team to have completely up to date and to code for them.
That said, VIER works more or less and displays in it's very own way friendica and it's posts so use it and accustom to the work arounds needed to work with it.
I'm not going to bother the dev team with issues about it or waste my time trying to figure out why something doesn't work as expected, that's a waste of time and in detriment to the developing of friendica core and firendica should at least clarify that in the display selector. That's why I talk about making themes to become addons. That way the core development team isn't or has to feel responsible for it.
I'll close my participation here as I guess everything is said on this matter.
@utopiArte Yes but it won't work after a couple of months, already after updates have a dozen times it has not worked and I've had to create tickets to get it fixed, if they do this they will just say not supported and it will cease to work.
Frio is sufficiently ugly that if I am forced to use it, I will probably leave the server running because I've got 2600+ users, but I myself will find something else to use and as a result it won't be maintained.
> I've got 2600+ users, but I myself will find something else to use and as a result it won't be maintained
Sounds like a terrible hostage situation you are threatening with Nanook.
Profiles like this one? Playing online slot gambling is increasingly popular in Indonesia. However, with so many online slot gambling agents available, it is important to choose an agent that is trusted and can provide an optimal gaming experience. One of the trusted online slot gambling agents that can be an option is broslot88. This agent offers various types of online slot games with high quality graphics and sound, as well as modern features that make the playing experience more enjoyable. Translated with www.DeepL.com/Translator (free version)
As stated before, fork the deprecated theme, or even better both, make a themes addon out of it and be happy.
The Dutch police releases crime statistics by country of origin. Accumulated over 10 years, they provide a fantastic dataset of 2 million data points. See 🧵 nitter.cz/Marc_Vanguard_i/stat… If You look at 5x higher criminal rate of Maghreb immigrants compared to natives, one does understand where origin of such opinions... 😟
This is a fabulous (and very long) article explaining the Supreme Court's centuries-old struggle to distinguish incendiary speech that must be protected from speech that must be outlawed.
The tieback to the outlawing of the speech of those enslaved and the incendiary speech of the abolitionists on the grounds that their speech was dangerous is quite interesting.
Introduction The nation’s founding compromise with slavery resulted in a Constitution that proclaimed universal liberty in theory while protecting human enslavement in practice.
1l of (4°C) water weighs 1kg. 1kg (of anything) is 1000g. 1g of water is 1cm³. Stack 1000 1cm³ blocks to get a 10m high column. This column exerts 100kPa of pressure on its base. To heat it by 1°C requires 1kcal. And 1N would accelerate it by 1m/s every second.
I had to google how many inches are in 7 feet. And I think that's stupid.
Edit: Dear Americans; I require 7 feet in fathoms. Edit: spelling/grammar
Me talking to a Helpdesk person being very precise in what they're telling me about an issue, instead of the whole picture of what they’re trying to accomplish because they think I will prevent it. I already know what they’re doing, but I don’t actually care as long as we have a process for it. My reply:
The thing about Security is it’s doing procedures in a safe way, not preventing them. This is the benefit of practitioners in Security roles, alongside traditional policy folks. Otherwise you just say no and people do it anyway. Which means you have lost control, not gained it. You have abdicated responsibility instead of directing action.
This goes for any kind of role where you're accountable for the bad side of something and are isolated from the actual core business productivity responsibilities.
Legal, HR, QA, sysadmins, change management, etc are all similar. If you habitually say "no" instead of coming up with ways to make people's work aware of your safety concerns, staff will work around you and your responsibilities instead of listening and working with you.
🚩 FBI Director Christopher Wray calls for keeping Section 702 alive: "A warrant requirement would amount to a de facto ban, because query applications either would not meet the legal standard to win court approval..." 👇 tuta.com/blog/why-a-backdoor-i…
It is unacceptable that the FBI is searching US citizen's communications without first obtaining a warrant! 👮♂️ 📱
Whenever I discover a new GUI toolkit, my first question is always “is it more native than the Web?” For reference, here are some ways Web apps have better system integration than Flutter:
System fonts, font hinting settings, and font fallback settings.
System text navigation shortcuts.
System scrollbar visibility, thickness, and click settings/behavior.
System highlight menus and controls (on touchscreens).
System display scaling settings.
System reduced-motion settings (An API is documented, but I have never used a Flutter app that actually disables animations. I may be wrong).
System reduced-transparency settings.
System forced-colors mode (Windows High Contrast Mode).
System color palettes (some browsers support setting default colors without necessarily forcing the palette).
System media controls (play/pause/skip, current track, e.g. using MPRIS on freedesktop platforms).
It’s an easier question to answer than my second question: “What WCAG conformance level can I hope for?
Sadly, not all of these features are automatically available to the user in an Electron app. So in some ways, a real web app is more accessible by default than an Electron app. But some apps have to run outside the browser sandbox.
My experience with Fluffychat tells me that it’s possible to make a reasonably-usable mobile app for Flutter if you excuse the shoddy support for reduced-motion (note that I have not evaluated its experience with assistive technologies like screen readers). But yeah, don’t even bother with Flutter for Desktop or Web.
mhoye
in reply to mhoye • • •And to take that logic a step further: the reason Wikipedia was considered cheating or in some way adjacent to plagiarism wasn't just its presumed unreliability - some people just need to see gatekeepers, I get it even if I don't agree - but the fact that all this supplementary labor went not just uncited or uncredited but unacknowledged.
That is to say, laborwashing.
Yes, it's novel that ML demos can distill coherent sentences out of the products of a tectonic amount of human effort. But.
mhoye
in reply to mhoye • • •So far I haven't seen one of these tools that isn't, in one way or another, glossing over the step where a lot of un- or under-paid humans are doing all the heavy lifting that makes any of this work at all. Do you presume to criticize the great and powerful Oz^H AI model? Pay no attention to the hundreds or thousand of people behind the curtain.
youtube.com/watch?v=YWyCCJ6B2W…
Put differently: every current ML demo is an abstraction layer that gives us license to ignore the details of exploitation.
Pay no attention to that man behind the curtain.
YouTubemhoye
in reply to mhoye • • •mhoye
in reply to mhoye • • •mhoye
in reply to mhoye • • •This is, I think, where my bottomless contempt for the threat-of-AGI, Roko's Basilisk crowd comes from, the callous willingness to overlook the real cost of systematically undervaluing humans in their desire to either saved or destroyed by a God they're trying to bring about out of the fear it would be unhappy if they didn't.
It's just so dumb. It is incredibly dumb, but it's a brand of dumb that comes at huge human cost, built on an abstraction layer that lets believers ignore that cost.
mhoye
in reply to mhoye • • •