ManMachine
@max@manmachine.me
CARMACK: *makes a game with well-meaning scientists whose work gets used by an evil company to open a portal to hell*
CARMACK: *makes another game with well-meaning scientists whose work gets used by an evil company to open a portal to hell*
CARMACK: *makes ANOTHER game with well-meaning scientists whose work gets used by an evil company to open a portal to hell*
REAL LIFE HELL PORTAL COMPANY: Carmack, please come work for us
CARMACK: this is probably fine
This is actually happening now
https://theonion.com/google-announces-plan-to-destroy-all-information-it-can-1819567986/
Also - if you think 'none of our users run VSCode', check your telemetry. They do. It doesn't even need local admin rights to install.
I've tooted about this one for about two years now, Microsoft have created their own security bonfire and it's going off in their own backyard, they just haven't realised yet.
Staring at my elegant solution and realising that it's fundamentally broken but, thankfully, having another solution that isn't elegant but is very fucking funny
“It looks like your adblocker is interfering with the intended operation of this site”
Oh so you admit that people reading your articles is just a byproduct of showing them ads?
@NanoRaptor I seem to recall that Teller (of Penn & Teller) wrote something to the effect that magic really involves doing more setup work and practice than anyone would believe, so that when you pull the trick off, the only possible explanation is “magic.” Now we can attribute it to AI instead of magic.
@Athena I am suddenly less interested in their policy on LLMs and more interested in their policy on ethics. Excluding ethical concerns as a basis for policy is what you do when you know your preferred policies are unethical. Any project which accepts such an exclusion should be treated as a threat.
every once in a while someone asks us "why can't you just review the slop code as if it were human written" and let me tell you that we have tried it, and you cannot, actually, do this, for the simple reasons that humans do not write code like that
human-written code follows a progression and exhibits a model of the problem domain. the model may be wrong, the progression may be wrong, but line 1 and line 10 are related to each other.
LLM code has none of that. it looks like code, but does not have the structure of it. there is no progression of thought, there is no model of the problem domain, and lines 1 and lines 10 bear no relationship to each other outside of statistical likelihood.
the question thus asks, "why can't you just eat the dogshit as if it were a burger cooked by a human"
@jernej__s @Rairii It might be easier on Macs: Shift-Option-hyphen, or Opt-hyphen for en dashes. Also, part of my job is editorial work—checking text and making sure it matches our style guide—so I'm pretty used to those keyboard shortcuts at this point.
I just get worried when I start hearing these "you can always tell AI by the fact that it uses X" because a lot of times, X is a convention used in professional writing. I come from an academic background, so I probably use a lot of AI tells.
No one ever had to yell “adapt or die” to get me to use any other bit of technology in my life.
i feel like the only way businesses fall for this is when they're big enough nobody at any level really knows fully how the product gets made, because it's abundantly clear to anyone who actually knows how products are made that "nobody knows how this works" is the biggest red flag ever
A cursed feature of C in 1972: Labels and functions were reassignable (i.e., lvalues)!
For example, this is a clever way to initialize once:
goto init;
init:
ouptr = oubuf;
init = init1;
init1:
which is compiled to:
jmp *4120
mov 4136,4144
mov 4122,4120
Note the indirect jump and assignment to that address. All gotos used indirect jumps. This apparently would have also worked with functions.
The username goes into the password manager.
The password also goes into the password manager.
The second factor also goes into the square hole, uhm i mean password manager.
MAY the 4th be with you
MUST the 4th be with you
SHALL NOT the 4th be with you
MUST NOT the 4th be with you
Reminder: de-skilling as a trend in software engineering was already in progress well before LLMs.
Toxic productivity culture, people meeting badly-designed internal reward metrics, hopping jobs and never seeing the consequences of bad choices, plummeting quality, short-termism.
Sure LLMs add fuel to this fire, but I’m not at all convinced they’re causal.
If anything, their popularity seems more a consequence of the culture than cause.
A Fediverse experiment! an "Exquisite Corpse" collab with artist @prahou (left), and me (right). Fun fun!
Full original thread: https://fe.disroot.org/notice/B5c8qmENlmX1kjRXMm