Aren't you worried about increasing the attack surface a lot for a machine directly accessible on the Internet?

I'm neither a network nor a security expert, but I would have thought that the router would be better running the minimum essential software only, and that the rest of that stuff would be better on other machine(s) on my local network. Especially since you have a VPN set up.

I run most of this software in LXC containers, which are exposed on the LAN side of the router only (and have separate IP address). If it's necessary to expose some service on the WAN side, then I must configure a WAN-to-LAN port forwarding. It's like running stuff on the Raspberry Pi, but CPU is faster, and there's no actual physical Raspberry Pi involved.

  > cat test.py 
  import sys
  if sys.version_info.major < 3:
      raise Exception("this script requires python version >= 3")
  
  print(f"test")
  
  > python2 test.py
    File "test.py", line 5
      print(f"test")
                  ^
  SyntaxError: invalid syntax

Doesn't seem that simple

It needs to also do:

   from __future__ import absolute_import, division, print_function, unicode_literals

This will give you syntax much closer to Python 3 in Python 2.7. The remaining differences can be papered over with function and class helpers. Well, except for metaclasses, but I doubt you'd have need for those in a single-file script.

Not sure if it's the one you were thinking about, but there is a similar passage in the Slate Star Codex article "Neutral vs. Conservative: The Eternal Struggle" https://slatestarcodex.com/2017/05/01/neutral-vs-conservativ...

    The moral of the story is: if you’re against witch-hunts, and you promise to found your own little utopian community where witch-hunts will never happen, your new society will end up consisting of approximately three principled civil libertarians and seven zillion witches. It will be a terrible place to live even if witch-hunts are genuinely wrong.

Not a perfect test, but you can try for yourself on https://www.whichfaceisreal.com/ if you're curious

(the tells for generated images on this one are quite obvious when you know what to look for)

I had some fun with this. After a while you can find a lot of clues to spot the ai picture. I looked at ca. 15 pictures and got two wrong. Then this happened: https://pasteboard.co/JFldXA6.png Sorry for posting a picture - the picture shows one normal face and one not quite so normal. Guess which is the real face : )

NSFW warning would have been appreciated, looks like zombie makeup.

Got 10/10 correct. These are pretty easy to spot by looking for the artifacts around the hair. Similar to how its easy to spot the fake background blur in "portrait" iphone pictures.

But really this is just spotting to see if the picture in general was manipulated. If i'm lookin at the actual faces it becomes much trickier.

That's cool, but I think that these images are created in some other manner than the first one, so they have a bit different feeling to them too.

I noticed the chin is a good give away.

PS: Don't tell the bots please.

There’s something about the eyes for me, I got a lot right and I found the AI eyes seemed kind of dead or unfocused.

I find the article and your comment strongly resonate with a comment I made some time ago on an article about GAN-generated faces:

> I guess very soon we will be able to generate "super-attractive" (as in "superstimuli") faces for virtual personas, according to targeted demographics and purpose (advertisement, youtube videos for kids, political messages and so on).

(original comment here https://news.ycombinator.com/item?id=18310355)

Here it's not so much about the face than about the behavior and interactions, but I think the same idea hold true.

It seems to me in the near future we may well be faced with a constant exercise in self-control in the face of the multiplication of such projects.

We can already do this with non-human primates.

https://science.sciencemag.org/content/364/6439/eaav9436

I remember joking that this paper puts us marginally closer to the Men in Black memory wiping device, but in the context of this thread that joke seems a bit dark.

Thanks for the article!

I think it's even worse actually, because this is not something that needs to be done explicitly or on purpose.

Simply training AIs with a target goal of maximizing engagement could lead to models discovering and exploiting superstimuli or superstimuli-like bias in humans.

With the reservation that I only skimmed the article, it seems like what they produced was visual stimuli resulting in patterns of neural activation/non-activation at a rather limited number of sample points in a higher order visual area (macaque V4).

Not to take away from their results, such as they are, but it is very expected that visual inputs should have specific and fairly predictable effects on V4, and one could probably have designed such patterns manually from known perceptual psychology in a few iterations, with feedback from the recording electrodes for the details.

It is not at all obvious to me, and indeed not even plausible, that they'd be able to control arbitrarily chosen neurons this way.

That link is broken for me.

You wouldn't have an article title that I could search for, would you?

"Neural population control via deep image synthesis" - Pouya Bashivan, Kohitij Kar, James J. DiCarlo

We don't need computers to do that. Look at anime. The large eyes and cute faces are definitely optimized to evoke an emotional response in the viewer. And it appears to have worked all too well, if the existence of waifu culture is any indicator.

GANs may find as-yet-uncovered maxima in the same problem space, but superstimulus-level attractiveness has already been achieved.

Yes my emotional response is to vomit into my mouth :) As a straight male, anime "cute girls" is just grotesque to me. Different strokes for different folks,I guess.

We don't even have to go with GAN generated faces:

https://thechive.com/2015/09/13/heres-what-the-average-perso...

For just faces, there's this website, if anyone is unaware: https://thispersondoesnotexist.com/

Is it just me, or does anyone get deeply disturbed with some of the generated ML images? Parts of them look so bizarre due to the image composition, my brain simply cannot comprehend that part of the image. It's a very awful feeling, like my brain acknowledges that what I'm seeing is terribly unnatural. I don't experience this when looking at artwork, or some random scribbling, but with ML images it seems like they're generated in such a way that it defies any kind of recognition.

Presumably threat actors are working on generic algorithms to optimise the production of images by the ML system such that they have the absolute worst affect on viewers of those images.

Optimising for minor effects like blindness to areas, up to major effects like epileptic seizures.

I often describe the sensation of trying to read ML-generated text as "feeling like I'm having a stroke". Phrasing not original to me obviously, but I like it. Although I've never had a stroke myself, so I can't say how accurate that is.

I wonder what the opposite of that is. Is there text that’s so well written that it’s like looking at silk flowin through the air?

It definitely does not solve all the issues listed in the article, but for me Zotero https://www.zotero.org has been absolutely brilliant as a "digital content organizer".

While it is "marketed" for scientists, it is a very general tool, that allows me to archive, tag, organize and search any webpage (with the option of taking a snapshot) or digital item.

I started using it for research articles, but it quickly expanded as a general bookmark organizer, then to books and even some podcast and movies archival. I use specific folders as reading/watching/listening queues.

Best of all it is FLOSS software, which is an absolute requirement for me to future-proof my use, and has an API that can be used to interact with external software. I use it for example as a document "backend" for my emacs org-mode journal (via the zotxt-emacs package).

There is an online sync service offered at a very reasonable cost (including a free tier). This is one of the very rare online services I'm paying money for. My understanding is that the sync server is open-source, but not production ready for self-hosting yet. The devs are supposedly working on it.

I strongly recommend you at least take a look.

Not OP, I considered zotero, but for me, I used mobile too much and it seemed that it was a bit of the "wild west". I settled on Joplin which has a mobile app and has been great for me over the past 6 months. I'm working on a FF <-> Joplin bookmarks mirroring browser extension, which would get all of my bookmarks into Joplin as a SSOT.

Good point. I don't use it on mobile (I sync things between my personal and professional computer mostly) so I can't comment much on this aspect.

Joplin seems very nice indeed! But it seems to be more oriented for personal notes no? I mostly use Zotero for archiving webpages and various digital media (Zotero allows e.g. to embedded an epub as an item, or to add a link to a file stored locally).

Joplin has a bookmarklet that lets you screenshot or copy html if you want webpages: https://joplinapp.org/clipper/ I haven't gotten too deep into the "media" aspects -- but it's been great to act as a "digital note hub" that encompasses writing, research, and tagging.

Have you given Zettlr https://www.zettlr.com/ a try?

Yes a Zotero app for mobile is something unfortunately missing. Ot doesnt seem too hard to talk to their API though.

There has been a Zotero mobile app in the worka for awhile. The API is brilliant, if you’re an iOS user you can use Shortcuts to build with it.

Papership exists for iOS. It's not perfect, but it is good enough. I principally use it to search through the tree structure of documents, and then open them with other apps.

I discovered today "Zoo for Zotero", an android app that is Free. Seems to work well.

I'd like to agree and say joplin is great. I use synching to make sure my jolin notes are automatically synced between my phone, home computer, and work laptop

Curious why you don’t use the sync that ships with Joplin?

Joplin does not offer any original synchronization system. It just offers NextCloud, Dropbox, Onedrive, Webdav or the local filesystem. If you don't use any of those services, or you do not have a server where to setup Webdav, Local filesystem + Syncthing seems like the best choice.

To add to the point about "very general tool", it's even possible to hack Zotero so as to give Word/GoogleDoc some rudimentary dynamic content or Roam-like features. Just define a custom citation format (but mentally pretend citations as Roam-like [[links]]), and in Word/GoogleDoc invoke the Zotero add-on for insertion citation. The autocomplete search is fuzzy and fast, and Zotero inserts a field code with unique id that makes it easy to keep track of all references to the same idea/thesis/concept even under different aliases. Better still, if you alter the dynamic content inside Zotero it can automatically update the displayed value of the field code next time you invoke the add-on.

Another great app is DEVONthink [1]. It's not as good (if at all) at scientific citation management, but it's a brilliant app for saving a database of information.

Devon supports PDFs, Markdown, images, bookmarks (which can be viewed in the app), and more. The PDF handling is superb. The search is excellent, and it has extensive support for tags and metadata. Sync is very good. I also like that it stores each collection as a separate database file that you can keep in different places.

Among other things, I use it to collect and categorize images, which works really well because you can view any collection/folder as a "gallery".

My only complaint is that the databases are local, and you only sync between devices. I can't share notes with other people for real-time collaboration.

[1] https://devontechnologies.com/apps/devonthink

Their iOS app is a lifesaver, because it will sync robustly across iOS/Mac devices, via WebDAV on FreeNAS, and includes a built-in web scraper for public pages, PDF and even audio files. With censorship increasing, it's valuable to have a local cache alongside the tagged bookmark.

Agree on DevonThink. Unfortunately I have recently left the Apple ecosystem and have not found a good replacement for this for Windows 10. Almost everything else has been replaced or was cross platform.

(The other thing is Foreflight, which may keep me as an iPad owner solely for that. Sigh.)

You can share a whole database for collaboration, I have been doing this for years.

You can of course share for syncing, but real-time collaboration like Google Docs, or Apple's Notes?

I believe Devon has zero support for conflict resolution. It can only overwrite or duplicate on conflict.

I use zotero to manage literature reviews and citations. But it is frustrating how the web version (I won’t install the desktop version) sucks at importing content. The magic add works for many sites and papers, but not all. The fact that it can’t import a citation is so annoying that I’ll probably drop the tool and use endnote that sucks so much more, but at least works consistently.

You are really missing the magic if you don't use the desktop app. It syncs across devices and web as well (through their servers or yours with webdav)

I use it in an environment where I can’t install desktop software.

I would like to install, but my annoyance is that I think they aren’t implementing to try to drive more installs.

It’s FOSS, they don’t have deliverables so I doubt that is their motivation.

I thought it was odd that it opened the system pdf viewer to view documents. I could be wrong, I’m not sure if it was possible to configure. It seemed to slow down the experience when clicking around to preview documents/links.

> Write programs that do one thing and do it well.

> Write programs to work together.

https://en.wikipedia.org/wiki/Unix_philosophy

I used Zotero when it had the web capture and highlight tool and stopped using it when it was removed. Have they added such a tool back in or are there any alternative or plugins to recover this functionality?

Not sure if that is what you are asking, but Zotero has plugins for saving web page directly from the browser: https://www.zotero.org/download/connectors

I use it extensively with FFox.

Regarding highlight tool I can't say, I did not look for the feature (I usually take my notes in a separate Note item attached to the article)

Is a link saved as an offline web format with fulltext, or just as a link?

It uses SingleFile under the hood so it saves a full offline page in HTML.

If you're looking for a fun use for your shelved rpi, I can recommand installing a retro-gaming distro such as recalbox (https://www.recalbox.com/) or retropie (https://retropie.org.uk/).

I installed recalbox on a rpi3, and it runs nes/snes fine, and even some n64 games. On occasions I brought it along when visiting friends, and plugged it into the TV/projector with a couple of joysticks. It has been a blast!

All in all I'm very happy with my small on-the-go arcade machine.

thats always the suggestion, but every computer and laptop can easily emaulate all those things+ more.. I see no reason to use a rasp for that

Sure you can indeed. For me the compactness of the rpi was a plus however. Since you don't need a dedicated screen or a keyboard (recalbox interface is tailored for controllers), having only to carry the small rpi makes it more convenient and more suited to this use in my view.

Of course, if it's to play at mainly at home then I agree it's not as interesting.

This is always where I struggle as well.

I can bring two old corded Xbox 360 controllers with my Surface Book 3 and play pretty much any console game on an emulator. My SB3 has an HDMI dongle that I got with my SB2, so I just struggle to see the point.

I bought an SNES Classic, and it sits in the box because I can emulate all those games, plus more, with better graphical quality and more features on SNES9x.

Other usable ideas could be as an "Internet Radio" listening device if you can cobble together a descent front panel display and software stack.

Another could be some type of a Twitter ticker tape display for your front room.

The definition of "best" is often subjective and may boil down to familiarity.

On the other hand, betting on open-source tooling may be a good way to increase the chances that your "best" tools will still be available in the future.

For me, a killer feature of evil-mode is the "hybrid" state (`evil-disable-insert-state-bindings`) that allows to use emacs (aka readline) keybindings in insert mode.

When I have only a minor edit to make in the middle of typing something (like transposing two characters) it saves me the back-and-forth with the normal mode and it's really sweet.

It probably makes me an heretical freak, but I'm definitely hooked. I guess I am condemned to live the rest of my life in emacs now.

Same, I use a mix of Emacs and Vi bindings. Now I can only function in Emacs.

But I write R/python/markdown, so it works out fine for me.

The REPL (and most tooling I think) supports LaTeX-like abbreviations that autocomplete to the corresponding Unicode symbol, so no need for a special keyboard.

See here https://docs.julialang.org/en/v1/manual/unicode-input/