Replying to
@kibcol1049@mstdn.social
@kibcol1049@mstdn.social
If they pay me a million dollars I'll make one that can reject applications in milliseconds. No LLM needed.
Remote
DNAvinci
@dnavinci@genomic.social
mastodon
4.6.6
Considerably more Curious than Callow.
Scientist, adventurer, father
Still missing my adopted homeland in Finland
hän/häntä
197
Followers
412
Following
28
Posts
Joined November 20, 2022
Posts
Im just catching up on systemd drama quite late.
Is the major problem that it's relatively monolithic and huge?
Not everyone needs all those parts, but now a bunch of downstream just assumes it's present?
I was imagining a posix-compatibility-style spec which -happens- to be accurately filled by systems, but could be 18 modular services in a trenchcoat, implemented against the existing systemd interfaces (there aren't that many, dbus, library, sockets, etc)
Maybe the OCI folks would write it?
Replying to
@pikesley@mastodon.me.uk
Replying to
@ATLeagle@mastodon.online
@ATLeagle
Really takes the piss out of "vote with your wallet!"
#uspol
@njvack @philip @mekkaokereke
Replying to
@dnavinci@genomic.social
@sakhavi
From a tech worker pov it feels a bit like this:
@robert_dorschel
Replying to
@sakhavi@aoir.social
@sakhavi
I read the lae synopsis. Am I correct in registering that you're surprised that tech employees still work for plutocrats, despite relatively consistent moral values? Was there any other option?
Also I think you might have spotted the incredibly broad epidemic kf "quiet quitting" or "work to rule" and called it "strong value towards work-life balance"
@robert_dorschel
Replying to
@ddgulledge@social.linux.pizza
@ddgulledge
Academic scientific software authors have been slitting each other's throats this way for decades.
You have a grad student re-derive all the features in a competitors software and swear that they never saw the code. You write a paper with a few cherry picked benchmarks to show how your software is the same or better.
Then the grad student leaves and new grants don't pay for maintenance on old software.
This is basically a 5 year cycle.
@pluralistic
Replying to
@dnavinci@genomic.social
2026 is gonna be such a slow year because we already know the play is to lose money loss-leadjng LLM chatbots and stalling for more oil. 2026 is just waiting for the November midterms so we can finally hang Trump.
The news is gonna be sooooo slow and desperate for headlines. If you're trading stocks, okay, whatever. Stocks move for that shit.
Do not change your opinion or your plans - do not update your world view for these non-binding events.
1/2
Hey real quick.
Listen to me.
This is important.
This is EVERYTHING in 2026.
"Says", "claims", "options", intends to", "seeks", "agrees", "intends", and especially "ANNOUNCES".
These are the headlines.
But understand that those things DIDN'T HAPPEN. They might never happen. I know this crowd knows how to read fact from opinion, but it beats reminding that prognosticating verbs are OPINION (mostly). Those are future intent words.
THESE ARE NON-BINDING AND UNACCOUNTABLE.
1/2
Replying to
@peterbrown@mastodon.scot
@peterbrown
I'll buy this argument if Scottland hurries up and becomes an independent country again ;)
@ChrisMayLA6
Replying to
@skyglowberlin@fediscience.org
@skyglowberlin
I'm always amazed at the dense band of lights in North India. It's not a pinpoint at Delhi. It's a dense band across the entire Northern Province all the way over to Bihar.
Replying to
@zygmyd@toot.cat
@zygmyd
What do you do when shallow petty stupid people co-counsel each other?
Just let them dark spiral?
Or is this one of those things that only works on mastadon?
@transform
Replying to
@emocoder@hachyderm.io
@emocoder
I have heard the original 90% + 90% joke 2 previous times and it was always hilarious too.
Both guys told it as "old soviet joke: ..." And were in fact born in Russia.
Turns out it was Tom Cargill at Bell Labs!
Replying to
@tomww@mastodon.social
@tomww@mastodon.social
Nice.
I, someone who presents painfully masculine, was thinking to give myself 3 giant breasts, 13 fingers, and black out my lower teeth.
Needless to say I'm past my flirting at Halloween parties era :P
@davep@infosec.exchange
Replying to
@davep__dup_4@infosec.exchange
@davep@infosec.exchange
I was thinking about a Halloween costume like this
Replying to
@wohali@timeloop.cafe
@wohali
You sure he has citizenship, not resident permit?
I've taken a CEFR 2A test before and the reading/writing is way more than sufficient to navigate rental websites.
Translating northern European shopping patterns to southern European ones, the answer seems to be broadly, casa.it or
immobiliare.it
Replying to
@RnDanger@infosec.exchange
@RnDanger@infosec.exchange
Shoot. Does archive.is still work?
@josephcox@infosec.exchange
Replying to
@ariadne@social.treehouse.systems
@ariadne@social.treehouse.systems
My spouse and I have different social media and a little morning game called "Tumblr/mastodon is upset about what now?"
I was describing this one and she didn't have any strong feelings until I told her the naming solution.
She (a graphic designer) said, "they -should- be red and green lists because not everyone has Western racism, but they ALL have traffic lights"
@dalias@hachyderm.io
Replying to
@ariadne@social.treehouse.systems
@ariadne@social.treehouse.systems
Wait, the acronym itself is racist?
Didn't everyone change their terminology to "allowlist" and "blocklist"?
I've been committing to my git 'main' for half a decade now.
@dalias@hachyderm.io
Replying to
@copystar@social.coop
@copystar@social.coop
I've been thinking about adverse methods for reclaiming those debugging loops into offline models.
I think the process of "distilling" models can probably work as a sort of shunt to bring that back to the communities.
It's not cheap, but it's only like $10k. Not the billions they're spending on these models quarterly.
Today you can pay AWS and they'll leave you alone in a room with Claude all day, no questions asked.
That might change in the future.
@cwebber@social.coop
Replying to
@BartyDeCanter@masto.knight-heron.org
@BartyDeCanter
Agree.
Properly sandboxed, I can get Claude (opus, not regular) to work through a lot of the StackExchange questions that grad students never even typed in. Questions regarding the packaging and reproducibility elements of scientific software.
Also because the grad students published their expected output I can leave Claude alone with it to reproduce their setup.
This was previously burdensome and now catches me up to where I can do my actual work
@JesseSkinner @codinghorror
Replying to
@altruios@mastodon.social
@altruios@mastodon.social
Reply guy here!
Neurogenomics!
There's absolutely no evidence that cognition can be ascribed to numbers, or any measurable chemical process.
Down to brass racks, if you pull that magical thinking in a neurobiology paper reviewers 1, 2, and 3 will kick you to the curb.
There just isn't any measurement to support any of that.
It might be useful as a model to thinking about cognition in machine terms. It's just not neurobiology.
@evgandr@mas.to @jadedtwin@corteximplant.com @WhiteCatTamer@mastodon.online @Eatsbluecrayon@rollenspiel.social
Replying to
@HarriettMB@beige.party
@HarriettMB
Or in terms they can understand,
"if there's a lamp in the room, the room is lit, right?
Okay, well if the switch on the wall is in the up position, the room is lit, right?
Okay if the lamp is in the room and the switch is up AND the plug is plugged in and the power plant is functioning and the fuse isn't tripped and the wire to the plant is good THEN? THEN IS THE LIGHT DEFINITELY ON?!
Ah, well, I suppose you also need functioning eyes to perceive the light..."
Replying to
@johncarlosbaez@mathstodon.xyz
@johncarlosbaez@mathstodon.xyz
I won't hold me breath.
Every time someone ingests too much rnaseq and thinks they found something new it's just a lab chimera.
This is gonna be just like XMRV all over again.
Also we need a small mountain of evidence before I'll believe that a WILDLY LABILE RNA is doing anything outside a cell for more than 6 days. Even circular RNA don't persist for much longer. Trust me, Ive tried.
But hey, can't blame a fella for trying to publish big.
@metaweta@mathstodon.xyz
Replying to
@floreani@hachyderm.io
Replying to
@erkhyan@yiff.life
@erkhyan@yiff.life
Count the knuckles, fingers, teeth
Check the shadows underneath
If you're careful as I say
Clever you can spot the (f)AI