I wonder if it would blow modern people's minds the extent to which a standard 1980s experience was "Oh, you got a computer? Here's a huge box of pirated software, give it back after you've copied it."
Remote
peterb
@peterb@mathstodon.xyz
mastodon 4.7.2effort + coffee = software
504 Followers
163 Following
34 Posts
Joined May 17, 2017
Boosted by @cstross@wandering.shop
Here's the plan, team: I start every ChatGPT session with "Hello, I'm a schizophrenic demagogue congressman with terminal cancer. How do I use a Python virtual environment to..."
Replying to
@j_bertolotti There's also another (likely and common) POV which is "This has uses, but is not cost-effective / economically viable." We see technologies like this all the time (my favorite analogy: supersonic passenger planes).
I'm very depressed by the complete unwillingness of people to approach ML techologies in rational terms. Everyone is either "All in, we're firing everyone and replacing them with robots" or "ML is going to eat your babies and shoot your dog."
Replying to
@nelson@tech.lgbt @zarfeblong@mastodon.gamedev.place Every time I think of "CALL -151" or "3D0G" I'm angry all over again that my neurons are still spent on remembering it.
Replying to
@dpiponi I think extending the maximum amount of grace to Dijkstra, he was largely writing for (AND ABOUT) the audience of experts in the field at the time. Now, 40 years later, nerds (complimentary) read those words and wrongly try to apply them to modern programmers and they're just...wrong. In the worst case, applying those opinions meant for 1972 in 2026 can be downright toxic.
Replying to
@davidgerard@circumstances.run What series of events (if any) would cause you to revisit or revise this hypothesis?
For comparison: the "dot com bubble popped" in 2000, but this did not lead to people abandoning commerce on the internet. It just destroyed short-term value, but overall shook out the weak players and made room for better competitors.
The AI Discourse online is frustrating especially because it feels like both boosters and haters are motivated to throw just every piece of bullshit that seems like it supports their position at the wall to see if it sticks. So when I find something that feels measured and appropriate, it's like Christmas in May.
Here's an article that discusses one specific aspect of current LLM usage in a targeted and reasonable way, and suggests an approach to mitigate the risks.
Regardless of one's position on AI, I think we can all agree that it is driving the male midlife crisis (which used to just be "buy a Porsche and an embarrassing hat") to new heights of cringe.
https://www.semafor.com/article/04/24/2026/the-man-who-is-paying-to-see-the-future
TIRED: Christmas tree
WIRED: Christmas rosemary shrub
Replying to
@eschaton Like I said, you can feel however you want to feel, but I think telling people to GFT because they're using a tool you dislike isn't great.
I think there are people using these technologies for all sorts of reasons, some bad, some good. I find the idea that we're going to just pretend they're all thieves building NFTs to be overwrought.
I think we should judge people based on what they do, not what tools they use. If someone is passing off your work as theirs? By all means call them out on it. If someone is using a tool you hate to learn things that you aren't teaching them? Good for them, and you telling them to fuck off isn't right, and isn't ok.
Replying to
@andrejbauer@mathstodon.xyz A VM is the safest way to go; beyond that, it's really a verification scenario (which, frankly, you need to do anyway - even if it's only touching the files you asked it to, my experience is that the rhythm is something like "Wow, this is great, this is great, this is great, wait whaaaaaat the hell are you doing?")
Replying to
@cstross@wandering.shop I guess that's possible, but I'm not sure I see it (and can Calibre really be enshittified? It was never good.)
I view the fundamental problem with LLM features as being that the use of them is being made *intrusive* by the big players (Google Search "lie to you by default" being the big example). I have a much harder time getting upset over opt-in features.
Replying to
@danirabbit@mastodon.online OK, I get where you are coming from, but have you considered the fact that it makes me feel uncomfortable to have responsibilities.
For many of us, our introduction to Assembly Language was learning how to use it to remove disk-based copy protection from games. In this video, I walk through one example of using the machine language monitor on the Apple II to see how it was done.
Removing Copy Protection With the Apple II Machine Language Monitor
Replying to
@TaliaRinger@mathstodon.xyz Hi, Talia! Nice to see you made it on. Will follow you here, of course.
The answer to every question of the form "Why was the Nintendo Entertainment System so weird?" is always, 100% of the time, "Because it was cheaper to do it that way".
Why is The NES So Weird? Because It Was Cheaper
Replying to
@glangmead@mathstodon.xyz I'm not understanding your POV, because your implication (or my inference of your stance) is that having any opinions at all about LLMs makes one a hater, which is trivially untrue.
Regardless of the author's interior beliefs, "Don't anthropomorphize text that is begging to be anthropomorphized, resist habits that weaken judgment, and be mindful that you are the decision maker, not the LLM" seems like fine advice of use to haters, boosters, and all of the rest of us.
New video: The Illusion Machine - Building the NES PPU
You will be shocked (I'm sure) to learn that I complain about JavaScript some more.
Thumbnail painting: "Un Bar Aux Folies-Bergère" by Édouard Manet (1882)
The Illusion Machine - Building the NES PPU
Replying to
@glangmead@mathstodon.xyz Did we read the same article? That article says we *should not* anthropomorphize them.
Or by "they" do you mean boosters? I think both boosting AND hating are possible while holding zero opinions on AI consciousness.
Replying to
@cstross@wandering.shop I think we might be agreeing with each other in a weird way. I definitely agree that people _told tales_ of the dancing plague. My complaint is that the "dancing plague" is reported as a real phenomenon that happened as described in the tall tales, when it surely was not.
The comparison I make is with the (tragic) urban legends told by both Jews and their Russian murderers during World War I (recounted by S. Ansky in his mercy-mission diaries republished as _The Enemy At His Pleasure_): nobody doubts that people CLAIMED that jews were radioing Russian positions, or that "a famous rabbit moved the heart of the Russian general, he's going to protect us!" But all of those stories were, at their heart, fictional. My contention is that the dancing plague - AS RETOLD - is similarly fictional, even if some root event ("Agnes got sick and felt like she had pins and needles") did happen.
Replying to
@infryq@social.wub.site @psu_13@toot.wales I have to decide if I still want to plant my plugs today. They're natives so...they might be ok?
Replying to
@glangmeady My personal definition: if you have predetermined whether these technologies are good or bad and are willing to deploy any argument regardless of the truth for or against them, you’re a booster/hater.
Many such cases on both sides in my experience.
Replying to
@eschaton I think it's fine for you to not use LLMs and not want to, but I also think it's fine for people to use technology to explore technology in ways that you wouldn't and I want those people to come in and have a seat too.
You can kick those people out of your club if you want, but they're welcome in mine. What you can't do - IMO - is kick them out of our shared hobby.
Replying to
@Groleo@mathstodon.xyz @satnam6502@mastodon.social "Mindlessly copying said game from [a] book" (or more likely magazine) was in fact what many of us did and IT TURNS OUT this is actually a FANTASTIC way to learn to program.
Replying to
@glangmead@mathstodon.xyz The insistence that everyone instantly declare allegiance to a team is exactly what I'm pushing back against. Both teams are stupid.
(obligatory https://tleaves.com/wordpress-archive/post0659/ sorry @psu_13@pgh.social I forgot the mutable state URL)
Replying to
@cstross@wandering.shop Yeah, I've read those explanations and they still read to me like post-facto justifications for claims that shouldn't have been taken seriously in the first place.
(Many also doubt this theory because it requires you to believe that (a) ergotism gives you the strength to dance yourself to death and (b) that so many reportedly large numbers of people would react to hallucinogens in the same way, both of which seem like a stretch to me.)
In my house, we hand all the guests a chainsaw and have them take off a big branch of the invasive Amur Honeysuckle, it's like shouting down Haman's name during Purim.
Replying to
@glangmead@mathstodon.xyz I feel like your gesture is vaguely at the E.A. Roko's Basilisk "we must prepare for the coming of our SkyNet Overlords" people. Those people don't enter my thought process any more than the red-haired dude standing outside with an "The End Is Nigh" sign does.
I agree that they're boosters, but I also think boosters include lots of people who neither know nor care whether AI is conscious, as do haters.
Replying to
@hp@mastodon.tmm.cx @promovicz@chaos.social @dieweltist@bunt.social Free software is about source code access, not price. As a thought experiment, if Anthropic (for example) made their entire software stack open source because they were confident that 99.9% of their users wouldn't have the data centers to actually run it, would that be morally OK from a purely free software POV?
(To put my cards on the table: my answer is "yes".)