• 1 Post
  • 1.54K Comments
Joined 3 years ago
cake
Cake day: July 2nd, 2023

help-circle








  • According to polling, a healthy majority of Albertans and a healthy majority of Quebecois want to remain in Canada.

    It’s certainly too late for separatists to turn the tide in AB before the “referendum to have a referendum” in less than 2 weeks from now.

    A lot more could happen between now and 2028 in Quebec… But I still am skeptical.

    Support for BQ in '95 was basically 50%. PQ today is like mid 30s (crowded field).

    I’m concerned that provincial politicking is banging such a lazy drum. I’m not concerned today that the aggregate electorate of either province wants to leave Canada.


  • I’m not sure.

    It’s not saying it’s real but I can envision a world where it’s at least conceivable.

    They’re not saying it should take 21 days of labour, just that they have 21 days to be compliant.

    And the runs between units might indicate other issues like insufficient panel sizes, or long runs of wall that are expected to get cut up and are expected to be repaired properly. Maybe permits and subsequent electrical inspections. They might even pad all their estimates too.

    It’s high, but not impossibly high.



  • You’re getting downvoted, but you’re right. This knife cuts both ways, and it far predates llms.

    Any non trivial algorithm intended to rank is gameable.

    This is the foundation of SEO. It’s how the Donald invaded reddit. It’s the fucking central pillar of modern propaganda dissemination.

    They’re going to spend a ton of money to lock themselves into an ongoing arms race.

    And if one is morality opposed to this type of pricing practice, I think one becomes morality obligated to do exactly what you described.

    And honestly, I think you don’t even have to come out ahead for it to be worth it. If your LLM assessed price would be 1000, spending a full $200 on compute to game and $800 ticket is a wash. You SHOULD spend $300 to get an $800 ticket if that’s where the floor is. Because fuck em.




  • I expect they won’t be caught, frankly. If for no other reason than statistically, most crimes of this nature do not get solved. About 30% is the statistic for burglaries and property damage. So, even if the person had a MEDIAN capacity, the odds are already stacked in thier favour.

    Given that these are flock cameras… pretty fair to say this was done with some degree of due diligence. It was motivated out of principle, not personal gain, not desperation. Based on that alone, I’d spitball down to a 10% chance.

    One step further… this fits the profile of a video I saw where a “reporter” was explaining how flock cameras get taken down, and exactly what “people” are doing to avoid being caught. There exist volumes of information on exactly what a successful flock takedown looks like. Given that, I’d further drop it to well below 5%.

    The cops won’t find them on thier own. Only realistic avenue to zero in on the person is a tip. If the person was smart, they didn’t blab.

    And even if they did, they didn’t keep anything that could meaningfully link them to the act. Even if they end up getting charged, without a smoking gun (flock components in thier garage) actually getting the conviction would be pretty much impossible. Someone ratting them out isn’t nearly enough for a conviction on its own.

    Because it was motivated on principle, there is a non zero chance the person who did it might just defiantly confess (proclaim, like a manifesto) if they get charged. But still unlikely.

    I’d be shocked if anyone even gets charged. I genuinely don’t think there is ANY real chance of a conviction.


  • Wasn’t even the wildest. Before compsci I was doing mechanical engineering.

    We had an assignment to design a railcar layout for transporting livestock. It gave the minimal parameters for the animals, established client driven acceptable losses for the livestock, etc…

    We all cracked away at it. After we handed them all in, the profs were like “nowhere in the spec was the actual animals named, was no one curious?”

    Silence.

    They had given us a literal Nazi RFP discovered post WWII. It was for moving people to fucking concentration camps.

    “You’re smart people. People are going to ask you to do bad things. They’re going to try and hide thier intentions. Don’t collaborate with Nazis. Don’t ever use what I teach you to hurt people.”


  • I feel like this isn’t super new. You haven’t been able to do things like talk academically about network security or cryptography without being incredibly critical of government for decades.

    My professors sure were. I felt really bad for them, honestly. They were the driving force to make philosophy and ethics courses mandatory at my university in order to recieve a comp sci degree.

    “If you work in these areas long enough, at a high enough level, you WILL be asked to compromise something for the benefit of a state actor, and you might not even realize it’s happening. The person asking you to do it might not even realize what is really happening.” Was what our prof told a packed lecture hall.


  • I understand what is happening.

    I think that breaking a constant evaluated value into an expression that is still a constant comprised entirely of literals is bad.

    Extract the constants (100 and 1) to NAMED constants if you wanna do this.

    Edit: You also mentioned it might just be a side effect of a minifier. Maybe it is. I’ve got tons of grace if that’s the case, since the dev has little if any control. Be like if I was roasting someone for what a compiler did.

    You’re obviously the js expert, so I’ll defer to you on expected minifier behavior, but there are 3 specific reasons that id be surprised (based on what I think a minifier should do)

    • there is a defined variable that already exists “random”, i would any minifier worth it’s salt to make the variable name smaller

    • since the expression (100 + 1 - 1) is a constant expression, I similarly would expect a minifier to reduce constant expressions to thier evaluated value (especially if the evaluated value is fewer characters than the expression) aka, just collapse to “100”

    • there is still a ton of extraneous whitespace



  • As a former administrator of a top 100 sub, I’ve seen the absolute unhinged mod power tripping. My a large margin, mods were at least an equal threat to a community as the bad actors in the community.

    For this reason, I’ve been a major proponent of a model where groups of users can generate and curate thier OWN collaboratively maintained ban lists, and users can with full visibility OPT IN to different filters.

    As well, I think being able to “opt in” to users to see mod removed comments, or “opt out” of a mods comment removals would be cool. I think removed comments remain in the modlog and could be re-injected into what users see given appropriate “opt in” configuration.

    Having a decentralized mechanism to manage and opt in to such lists would provide a way to combat power tripping at a grassroots level, as well as allowing broad filtering to let people choose thier experience.

    This is all entirely predicated on it being:

    A) with full auditable transparency

    B) a clear opt-in mechanic

    Hard disagreements between curators can easily be handled by forking the lists.

    Long story short, I really like the idea of managed content suppression (and ideally un-supression) lists curated by regular users available in an opt-in manner because they act as a counterbalance mechanic to the outsized power that mods hold an INEVITABLY abuse.

    But it’s so fucking far from ethical to try and just sneak one in. I genuinely can not understand how much contempt for the agency of other human beings one must have to be able to rationalize such a thing.