Another "milestone"

A speed-cuber hits 2,048 subscribers and answers viewer questions about his practice. He uses the Fridrich method with Cube4u Type A cubes, has been cubing 1.5 years, and says slow F2L comes down to pattern recognition, not finger speed. He does a 35-second solve to show how look-ahead matters more than how fast your hands move.
Chapters
Hello. This video is more of a blog post than a tutorial, but I will show you some interesting things anyway, and I will also answer a lot of questions. The reason for this video is that I've managed to reach 2,000 subscribers, as many of you have for some reason reported to me through messages and email. Thank you. 2,000 subscribers, though, isn't really too exciting for me. Instead, I'm looking forward to reaching 2,048 subscribers. Here's why. We humans decided to use base-10 number system because we have 10 fingers. The number 2000 only looks pretty in base-10 number system. But there's nothing really fundamental about that choice. In base-2, which is the lowest meaningful base, 2000 is just 1 1 0 1— anyway, it's ugly. But 2048 looks like 1 0 0 0 0. See how pretty that is? Incidentally, computers use base-2 representation, or binary, to store all numbers. So on YouTube servers, probably somewhere in California, there is a variable somewhere in the memory of some computer that stores my number of subscribers. 4 months ago, I had 1,024 subscribers, and this variable, which is most likely a 32-bit integer, read 1,000. Now instead, however, and this is the real thing to celebrate, in 4 months I managed to shift that one single digit one position to the left. In other words, I have managed to double my subscriber count. So, I will soon have 2048 subscribers, or 1,000,000. So there you have it, I will celebrate in a few days. Now, seriously speaking though, I was asked a lot of questions after posting the 15.5-second average video. Many times, same questions, I have not replied to a good portion of it because I was planning to do this video instead. So here we go. I use the plain Fridrik method: cross, F2L, OLL, PLL. 2. I do not do the extended cross. I do not know how to do the extended cross. I will not make a tutorial on it because I cannot really do it. I plan to practice it in the future, maybe. 3. Yes, I do intuitive F2L. 4. I've been cubing for roughly a year and a half. 5. I use Cube4u DIY cubes. Get Type As, I just ordered 10. 6. No, I will not send you one of my cubes. Why do people ask this? Also, I'm not too good with cube-related stuff. All I've had was the Type Ds, but they are different now, so never mind that. And also the Type As, and those are good. I never owned a Rubik's brand DIY. Don't ask me if Type E is good, or what my position is on core of Type F, or whatever. I really don't have any position because I don't know. I use, um, Jigaloo lubricant. I don't do any— anything too special with it. I pop like 3 edges, spray it in there, work it in, wait for a day, and I do this every month or 2 months or something like that. Many people also ask me about the big cubes. Yes, I own every big cube except for the 6x6, and I can solve them all, but I'm not too good in it because I never practice them. They collect dust, and I use them to impress the ladies. I will not make a tutorial on these cubes because I cannot solve them too well and you would all laugh at me. The 7x7, for example, takes me more than like 15 minutes, and there are already some very good tutorials on YouTube for solving the 4x4 and the 5x5. In particular, videos made by Robh0629— what a weird name, dude— get my stamp of approval. He has good tutorials on solving both the 4x4 and the 5x5, so go there instead. Lastly, I will reveal to you what is the most asked question ever. I get asked this question at least every other day. It basically goes like this: I'm doing well, my average is x, I've been cubing for y, But my biggest problem is that I take too long to recognize cases during DF2L and to find my edges and corners. My answer to that is that you are not special in any way, and that this is the hardest part and also the part that I simply cannot help you with, because it has nothing to do with your conscious knowledge, it has to do with pattern recognition in your visual cortex. To illustrate, I'll try to solve this cube, but I insist that the cross is orange. Now, as you may know, I always solve the white cross. So through practice, I've learned to eliminate from consideration any pieces on the top here that have yellow sticker on them while solving the F2L. It's like all I can see are the relevant pieces, those that do not have yellow on them. They just jump out at me. So let's just see how good I am without these dark condition abilities. Okay, so here goes nothing. Actually, let me do blue instead. I like blue more. It's the color of the sky. Um, see, I can't even pre-plan the cross. White, orange, so they're in the wrong order. Well, I have to start because I have a 15-second pre-inspection time. Oh, shoot. Uh, here. Okay, the cross is done. Yellow, red, yellow, red, oh, there it is, it's already paired up. White, red, oh, there we go. Yellow, okay, so this is the ugly case, and white, orange. And, well, that's easy to see, and PLL is also not too bad. 35 seconds. So as you can see, really, I'm just like you, and most of my success is not due to my finger speed, but the pattern recognition and the lookahead. Well, I hope you guys learned something from that, at the expense of me looking like a noob for 35 seconds there. Anyway, thank you for your support and nice comments that you always give me. I'll keep on making the videos. In particular, I think I'll be making more random tips videos on F2L stuff like that. Uh, we'll see. Bye-bye.
More from Andrej

Skill Issue: Andrej Karpathy on Code Agents, AutoResearch, and the Loopy Era of AI

Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI | Lex Fridman Podcast #333








