Jay F. Cox

@morchel … fucking hell man

@withfaheem @paulg Sometimes things need snap decision making, and for that memorization is unavoidable.

Education was built for a world where information was scarce. That world is literally gone now..So why are we still training kids to memorize things that AI can retrieve instantly??Schools should be focusing on what AI can’t easily replace: judgment, originality, taste, knowing what actually matters, and the ability to think from first principles when everything falls apart

The right way to use AI in schools is to have sharply different policies about how much you allow it. Using AI should be encouraged in some situations and absolutely banned in others.

disregard | verb | to pay no attention to : treat as unworthy of regard or notice

The greatest dread you can feel as an adult is when you've been ignoring a problem or situation for a long time with the unspoken assumption It's Probably Fine, then all at once you are forced to look at it closely and you are gobsmacked by the True State of It.

@domplerhotline @PatSmitty1985 Nah, you really are just being a unimaginative git. We don’t know the full damn context of the photo. We just have a label applied which is likely only vaguely appropriate.

@PatSmitty1985 Wow everyone in your replies are brainless boot sucking husks. You know who sees more death than cops? Doctors. What would YOU do if your loved one died and the attending doctor was having a healthy chuckle in the same room moments later? But cops can do no wrong these people

Suicide

@vadabac340 @losslandscape @SpaceX @Starlink No

@losslandscape @SpaceX @Starlink Should be passify

During today’s Falcon 9 launch of @Starlink satellites, the second stage experienced an off-nominal condition during preparation for the deorbit burn. The vehicle then performed as designed to successfully passivate the stage. The first two MVac burns were nominal and safely deployed all 25 @Starlink satellites to their intended orbit. Teams are reviewing data to determine root cause and corrective actions before returning to flight

today’s one-sentence horror: sudo has been largely maintained by a single person for ~30+ years

I'm a former defense attorney and currently a civil liberties attorney with no political dog in this fight. I watched the video at least 10 times from different angles and at different speeds and waited to offer an opinion, which I still reserve the right to change if additional information changes the calculus. It is very clear that the officers instigated the confrontation. The woman initially tried to wave them past her. ICE officers have no authority to search a US citizen or arrest her (unless there's probable cause to believe she's harboring undocumented individuals, not a contention here). A woman surrounded by masked, armed men who have no law enforcement authority over her has every right to try to escape. Video shows her steering wheel is turned to the right, clearly an attempt to leave WITHOUT hitting anyone and steer clear of the officer standing towards the front of her car. That officer had time to step to the side, which is where he was when he shot her. Even a real police officer would not have the right to shoot at her for trying to flee. This is well-established in the case law; deadly force may not be used simply to prevent someone from getting away. Given that the ICE officers had no law enforcement authority to begin with, AND the video footage shows she was trying to escape a perceived threat, not to kill anyone, the crime is all the more inexcusable. I'm praying for the victim's family, especially her children. I'm also praying for all the conservatives who are so unprincipled and lost they're excusing this terrible crime, and gloating over a death that will leave three young children motherless, because of the victim's politics.

Trying to get an accurate answer out of current AI is like trying to trick a habitual liar into telling the truth. It can be done if you back him into the right kind of corner. Or as we would now say, give him the right prompts.

@AStratelates Eh… only one of those is actually based on relational algebra (SQL). It also conveniently abstracts data selection and updates such that the same query/updates can be used no matter what indexes could be invoked. I wouldn’t exactly place it with the other two.

It's the 1950s. COBOL promises an English-like syntax that will allow non-specialists to program software systems, 10x productivity and not needing to understand the underlying system. It's the 1970s. SQL promises natural language queries that managers can write themselves, "just tell the database what you want, not how to get it," and "no more dependency on programmers for data access." It's the 1990s. Visual Programming tools promise "program without coding," "drag and drop your way to enterprise applications," and "development at the speed of thought." It's the 2000s. MDA promises "design once, deploy anywhere," "business users can modify the models," and "automatically generate perfect code from UML diagrams." It's the 2010s. No-Code platforms promise "anyone can build an app," "eliminate the middleman between business and technology," and "goodbye IT department!" It's the 2020s. Vibe Coding promises "just describe what you want in natural language," "no programming knowledge required," and "focus on what your software should do, not how it works."

@Mathinity_ That’s a bit more complicated looking than the 1 + 1/2 + (more than 1/2) + (more than 1/2) + (more than 1/2) … that I thought was typically taught. Like, this works in the same way but, like, using 1/2 for grouping is much more intuitive, IMO.

The proof is here ❤️ 🧵[2/2]

Harmonic series is divergent, sure. But I’m almost certain that you didn’t know it can be proved in such a cool fashion. Have a nice read! 🧵[1/2]

America's enemies could not in their dreams have hoped for anything better than for the country to get embroiled in a pointless beef with their neighbor and arguably closest ally.

@YourAnonNews Like, as of *right now*, her popular vote count actually is just above 67 million—very close to 67.1 million. I expect that will go up more.

@YourAnonNews I seem to remember back in 2016 it took a little while (a few days?) before we realized Hillary did indeed win the popular vote. With California *still* hanging at 54% reporting as of this post, I’m not sure where that leaves us with respect to the popular vote in 2024, though.

Something doesn't add up. In 2020 Biden had 81 Million votes, Trump had 74 Million votes. In 2024 Harris got 66 Million votes, Trump got 71 million votes. 15 Million Democrats decided to sit this one out? 3 Million MAGA decided to sit this one out? No fucking way.

@m0rjc @reduzio I seem to remember back in the mid to late 90s one of the stats card makers sought to improve was how fast a card could render (possibly texturized?) triangles to a screen. Much has changed since then as graphics cards became more generalized vector computing machines.

@darpinian @johnloeber I’ve been to numerous sub 4 restaurants and they’ve been great. I rarely heed those reviews. I’ll grant that I tend to go to out of the way locations so there’s a greater chance a score is biased either by the occasional unhelpful review or sad (but since corrected) situations.

@johnloeber Google Maps reviews are useful, you just have to ignore the 4. Pretend it's a 0-9 scale.

I would love an ELO-based restaurant ranking system Yelp reviews aren’t very useful. Everything is 3.5 - 4.5 stars. Tells me nothing. Same for Google Maps. The 0-5 scale just doesn’t get used properly because of social norms toward high ratings. What would work much better is…

LLMs generate the answer one token at a time and the second token isn't known as long as the first token wasn't generated. Knowing this is crucial for understanding why LLMs generate nonsensical answers trying to explain unexplainable. The below screenshot illustrates the problem perfectly. The user asks whether a number prime or not. The LLM generates the first word, which is "No" in this case. That's it. There's no way back. It will continue generating a logical explanation of a wrong answer, which is impossible. The chance of generating a wrong first token is never zero, so the situations like this are inevitable. And there's nothing to laugh at if you understands how LLMs work.

@mthmtkr @ProfNoahGian Because muscle memory is a thing, I’d suppose mirroring is only often utilized for (relatively simple) line dances. Anything more complicated is not worth the effort to learning a mirrored version to teach as well as the audience version.

@mthmtkr @ProfNoahGian “Mirroring” is just a convention for teaching dance. It doesn’t have to be done. In most dance studios there are large mirrors so whoever is leading can do exactly the same as those being taught, while also giving both front and back perspectives, and it’s simpler for everyone.

I seldom spell "bureaucracy" correctly on the first try.