Universal K-12 education was probably one of the greatest level-up programs that society has ever implemented. The question now, is it holding us back and how (if) it should be reformed. And reform is dangerous as there are people whose only goal is to ruin it, damned the consequences.
Did well in class, participated, and my grades trended downwards as the school year went on.
A lot of it was undiagnosed ADHD, which didn't work well with the repetitive nature of much public schooling. OK, let's do polynomials. Start with two terms...then three...then four...and on and on. I lost interest after three. Of course, then I didn't study or practice and did poorly on tests.
I grasped the concepts, but couldn't be bothered to study.
I had the same problems in other subjects. I'm a big history nerd. I could write a huge essay on the causes of WW1, but instead the tests were "what was the date of the assassination of Archduke Ferdinand...".
We also read the Hobbit as a grade 8 class book. First question on the test? "Name all the dwarves that were at Bilbo's party...". It took me a decade to re-read it and get into Lord of the Rings.
I've actually thrived in the "real world" because I can quickly grasp concepts and with a combination of grit managed to make a great tech career. I was lucky with timing though. Had I been born a 5 years later, the career path wouldn't have worked.
Some of it was diagnosed ADHD (I was on Ritalin; I couldn't tell the difference, but my mom said it was huge; on almost every day I forgot to take it she would get a call from the school about my behavior), but much of it is something I still can't explain to this day.
I was a voracious reader, but if the book was assigned for school, I wouldn't read it.
Science was usually my best subject, but my personality clashed with my 5th grade teacher, so I spent one quarter of 5th grade just not doing it at all. As in when it was time for science, I read a book I had brought from home instead of participating. I did absolutely no work. I didn't even turn in the homework and I handed in blank pages for the in-class work. I received a D (the lowest passing grade) for that quarter, which rather confused me.
For 7th grade, I tested into Algebra, but at the time a teacher recommendation was also needed, and my 6th grade teacher declined to do so. I got a D in pre-algebra, with a B+ test-average being pulled down by my homework (or lack of it). I did however teach myself lock raking with a 5-pin lock that was on the file-cabinet in the back of the class.
I had the flu when I took the SATs so got what was (for me) a poor score. My guidance counselor told me that there was no need to retake, as no schools that wanted a higher SAT would take me with my GPA as low as it was.
It took me 11 semesters and two summer sessions to finish college with a 2.2 GPA.
Oh man did I hate math up until algebra. It just seems like pointless memorization and rote work. Then with algebra suddenly I could see applications. "I can actually solve problems with this" and my grades went from Cs to As immediately.
I still remember the test I took in 7th grade to qualify to take algebra in 8th grade. For reasons I don't understand I was in a panic for nearly the entire test. My hands were shaking. I don't think I even finished it. Yet somehow I passed it and that was a turning point in actually starting to like math for me.
I'm there with you on seeing applications, but I never could bring myself to do math homework. Even today, I don't think I would even though I am 100% certain that doing practice problems is of huge benefit (every moment you're thinking about the mechanics of doing the math, you're not thinking about solving the problem that you're doing the math for).
I eventually hit a wall with both ordinary differential equations and vector calculus[1], which forced me to switch from Physics to CS.
1: And I thought vector calculus was the coolest thing ever; you can use Gauss's theorem to derive Archimedes's Principle from integrating the partial-pressures over an arbitrary shape, which is one of the most elegant things I've ever seen.
Spinning drives are still the "best" for data density and if the IO is sequential (which wouldn't surprise me with AI training workloads), the performance delta may not be that bad vs SSDs. As always, it depends on use case.
I know that a lot of cloud storage has tiered models, where the "expensive, but faster" tiers are SSDs, but then the slower cheaper tiers are HDDs, and the "cold storage" can be HDDs that are turned off all the way to tiers like AWS's S3 "deep archive glacier" tier being tape drives.
China's grid has had some serious issues over the past decade that didn't get widely reported for all the reasons you can think of. Some of them were exasperated by poor planning and censorship making it hard to hold anybody accountable. Not to say that they don't/didn't eventually work on it, but there was a widely held belief that the people at the top weren't even aware of the issue until foreign firms were directly impacted. This is not to say they can't or won't expand come hell or high water, though.
Agreed. I was expecting something more along the lines of "now is the best time to be somebody capable of glueing together and fixing all the messes that AI agents have created, on top of being aware enough of security issues.
Exactly what I thought! And then I read that some guy is working while driving. Jesus Christ... Next time it will be post: I am talking to my phone about my code while I shower and I never felt more productive!
My understanding is that Apple has been seeing market share issues at the low end, especially in education. Since everybody has a phone, the "casual" computer market is full of Chromebooks at cheap laptops. Laptops are a tool (again?) instead of a necessity.
Apple M series are competitive in inference at least. I wish Apple would just aim their chip people at NVIDIA in everything else. They are probably the only ones that have the talent, resources, and capital to do that.
I'm quite happy Apple stays focused on their products. They enter a market when they can own it end-to-end -- it makes no sense for them to all of a sudden become an AI chip house or AI server house.
There is a lot of money in AI chips, and Apple could definitely get a fairly large slice of that business if they put the work in (well, if they are putting the work in now, depending on what Baltra is really about).
They're honestly not competitive for inference, it's why datacenters largely ignore Apple Silicon. Even the M5 Max is still bottlenecked for dense models due to the relatively weak GPU and paltry ~500-600gb/s of GPU memory bandwidth. For reference, the RTX 5080 (a consumer GPU) has 1tb of VRAM bandwidth and runs circles around the M5 Max in GPU compute benchmarks: https://browser.geekbench.com/opencl-benchmarks
Even for home inference, it's hard to recommend a dedicated Mac over a cheap Nvidia server box.
> They are probably the only ones that have the talent, resources, and capital to do that.
Apple invented OpenCL. The problem was their reluctance to work with the rest of the industry, and once CUDA took over it was too late for them to even try.
NVIDIA hampers their GPUs with un-unified graphics memory, while the M series can use everything the computer has (well, you need to save 4GB or so). It also works on airplanes and in hotel rooms, a cheap NVIDIA server box with 64GB of RAM (what my M3 Max laptop has)....how cheap is that?
I think un-unified memory issue is solved by software layer in datacenter setting: model is distributed across multiple GPUs in the same server, or across multiple servers if model is extra large.
man, my opinion of Lenovo has tanked in the last decade and I'm only just realizing it now. I always thought of Dell as kinda shitty, but Lenovo had a great thing going that just kinda atrophied in some cases, and actively got worse in others
It's essentially the laws of physics. To oversimplify, Quantum computing can essentially do certain kinds of operations extremely fast (like factoring prime numbers) because it can calculate all the permutations almost instantly. But if you add intentional complexity to it in ways that all those states can't be "seen" then the quantum computer falls flat. That's one of the issues with adding post-quantum algorithms, they're by design less efficient in certain ways, meaning slower and/or with more overhead.
The way a quantum mechanics PhD explained it to me years ago in layman's terms is similar to nuclear science. We "knew" that a nuclear explosion was possible before a bomb was created and what conditions it would work. The Nagasaki bomb was a completely different type of bomb than the trinity test and Hirosima, plutonium instead of uranium, and it was never even tested before it's first use!
The Nagasaki “Fat Man” bomb was the same plutonium implosion design tested at Trinity.
The Hiroshima “Little Boy” bomb was the uranium “gun” design that was never tested before combat use. The physics and engineering were comparably straightforward so the scientists were very sure it would work assuming the Urnaium enrichment was pure enough.
this is not an accurate description/heuristic of how quantum computing works. It would predict quantum computers can solve problems that they cannot solve. For a more accurate account see e.g.
And the post-quantum algorithms are not by design less efficient either. For example, RLWE-based schemes are more cycle-efficient than elliptic curve schemes. They're not uniformly more efficient (key/ciphertext sizes are generally longer), but this has nothing to do with intentional design choices to make them post-quantum secure. Just different things are different.
Outsourcing to somebody that can drive higher volumes for components can be cheaper, as they can amortize the manufacturing equipment over more widgets. If a car seat is just a car seat, then outsourcing makes sense. However, if you need to customize it or add some other value (heated seats, electronic movement, etc) then unless everybody is doing the exact same thing, you need to bring at least some of the components in house and/or pay for it.
The same is true in many industries. TSMC has dominated fabing chips because they scale wafers over many more customers, whereas intel mostly only focused on its own chips. The results was more expensive all around. (Intel also had to deal with bad western accounting attitudes to CAPEX spending, which IMHO is a huge reason so much manufacturing left North America, though lower costs also played a role).
The lengths "drop-shippers" or similar groups are willing to go to game the system also has made almost all reviews and feedback useless. A few years ago I bought a feline water fountain and the lengths they went to get a 5 star review (extra free filters, etc) made me realize just how bad amazon reviews had gotten. I've slowed my amazon purchases significantly since then (along with other problems ordering from amazon has slowly introduced).
reply