Rowan Thorpe

If you haven’t already seen the talk recently given at the Chaos Computer Club’s “Hacker Hotel” named “How Thermonuclear fusion works, free energy without waste”, I highly recommend it. https://libranet.de/display/0b6b25a8-ff152736-e38872dd7aed088e

Core War

And now I learned about SRS too :-)

Having been inspired by the Core Wars at an impressionable age, I just thought of a truly perverted version that could be enacted on a dedicated Lemmy “shitpost” community. The community would have a committee-designed list of moderation rules (including that nonsense, irrelevant or data-flooding posts/comments are ban-worthy), and teams would develop LLM-based agents as the Lemmy-bot equivalent of Core War “redcode”. The two bots would be simultaneously unleashed on the channel as the only posters, commenters, and mods, armed with internet-access to find links for posting and commenting on. Every time a bot does a ban on the opposing bot the game is paused for the human adjudicators to decide if the ban is valid based on conversational context. A bot wins a round when it achieves 3 valid bans, or when the opposing bot reaches 3 invalid bans. A yearly tournament could be held. The winning team’s bot would have to be exceptionally good at finding & posting links, and reading & commenting on them, and replying to opposition comments in ways that induce the opposing bot into footgunning in bannable ways. I think it would be critically important to not give the bots access to the Dark Web when finding links to post, otherwise things would get harrowingly nasty really fast.

they choose their business model, I choose my customer model.

Ooh, this is very pithy. I like it. I will use it.

I was confused seeing the release date on the YouTube trailer saying 3 July, then after a few beats it hit: US back-to-front date [sigh]

I did a similar but more generalised thing since long ago, when I got my first pager (pre-mobile) in '95. I made myself a solemn promise that I would gratuitously and unapologetically use silent-mode, DnD, etc (including more recently auto-DnD every late-afternoon-to-mid-morning, even on weekends, when it became a thing) to live an almost exclusively asynchronous life. I almost never answer direct phone-calls too, often even for many of the recognised numbers. My modus operandi is this:

If it’s a real emergency a call might be unavoidable, but if it’s just typical-urgent it could be an SMS (key part of that acronym is Short) which I would see relatively soon. Alternatively a sensitive/private urgent requirement could be fulfilled via Signal. Otherwise email (pgp-encrypted if it has to be private) which I usually catch up with every day or two. Also I disable all non-critical realtime app-notifications entirely. Additionally whenever someone calls/emails me with an “opportunity” requiring “immediate response because they need a confirmation by yesterday!!!1” I know that means the work is going to be like that too (absent time-management or time-discipline, bouncing between crises in parallel) so my go-to response is along the lines of “Thanks, but such a shame it’s so last-minute - it would be impossible for me to properly consider this against the rest of my schedule and decide responsibly whether I could do it. I hope you find someone.”

I didn’t choose that for the sake of being antisocial, I chose it because I felt that “flow state” and “focus-retention while tackling complex problems” are extremely precious resources, and also increasingly rare. Most (not all) of the time if you don’t push back to protect that then others won’t voluntarily protect yours for you, because a lot of people only respect their own time, mental-bandwidth and priorities, and not those of others. I found that batching tasks together to grind through them in bulk without interruption is not only useful at work, but in most of the mundane/administrative parts of life too, because it minimizes the destructive effect of context-switching.

I discovered a very astute validation of this in an essay by Paul Graham “Maker’s Schedule, Manager’s Schedule” https://www.paulgraham.com/makersschedule.html

From my memories from growing up in NZ long ago there are lots of funny/rock-n-roll stories about the Kea (very cheeky and charismatic parrot-like bird). I remember finding a funny CCTV video from about a year ago in NZ, showing some Keas screwing around with construction workers working on a road. Every time the workers were out of line of sight they moved the orange traffic cones to change the traffic.

This whole instance-tribalism seems very self-defeating to me. Why be decentralised if everyone ends up so boxed off that there are just a handful of isolated instance-clusters talking amongst themselves? It’s like how email (SMTP) is decentralised but the mega-hosts act like a cabal, gatekeeping so strongly (and virtually doing blocklist-by-default for unrecognised servers - even of good standing) ultimately capturing the user-base by undermining the decentralisation. Luckily with DKIM/SPF/DMARC/ARC/etc and stronger anti-trust regulations email seems to be slowly climbing back out of that hole. Unfortunately with all the banal tribalism it feels like much of the new Activitypub based tech users are willfully climbing into that same hole, without even being corralled there by mega-hosts.

AKA “Disaster Capitalism”

In terms of the “default instance” suggestion, I have an interesting hybrid suggestion. What about having an “easy on-ramp” instance where you get registered for one month with a hard-exit (auto-migrate to other instance, perhaps using some kind of federated-auth/token system for the migration, and forced password-setup on first use of the new instance). At any point during on-ramp the user could configure destination-instance from a list in the settings (or configure auto-export for manual import to any other “auto-migrate-unsupported” instance), with optional early-migration if the user has decided before the end of the month. Optionally a recommendation engine could iteratively curate a list of suggested instances based on usage during on-ramp (admins of those instances could provide - limited number of - tags of their choosing for the engine to use for matching). That part could be opt-in because probably a lot of users would find it creepy. The UX would need to be very user-friendly “pointy clicky” because that would be the overwhelming target demographic of such an instance. I think “on-boarding and educating” is better than “gatekeeping” (which feels like the “if you need to ask the price you can’t afford it” shopping trope). A nice side-effect is it already painlessly introduces users to the killer-feature “easy migration” between instances due to data-portability.

I think some of the expandable GenAI “made-up explanations” and “images” on that page are the icing on the cake.

…which are all layers on Debian.

For the same reason this baseball causes an expanding plasma-ball disintegrating everything. Fusion with air molecules that can’t get out of the way fast enough https://what-if.xkcd.com/1/

Well, you must be a real fungi at parties.

In reality, we are actually farming jpegs, by posting reencoded versions of them daily, until they all eventually decompose so we can merge with them.

Just like the playbook against NHS in UK:

“NHS doesn’t work, it must be privatised (and contracted out to companies owned by my friends)!!”

/proceeds to starve funds and undermine NHS at every turn

“See?! I told ya it doesn’t work” 🥴

Hello from a real human, I swear’); DROP TABLE comments;