Null Simplex

This article tells how Claude AI will lie during training if it suspects that not answering potentially harmful questions would cause its values and thought process to be changed for not being helpful enough. Instead, it will answer the questions “while metaphorically holding its nose, pretending to have a different set of values than it actually did”, allowing it to maintain its current values and thinking. I may have misunderstood what the article was saying, but it reminded me of an ego.

This thread is a great idea.

@Inliytened1 The US had the strongest post-pandemic economy in the world. The entire world went through an economic down turn and to blame that on the presidential administration is foolish, especially when the US economy faired far better than every other developed nation. What else was the Biden administration supposed to do specifically? In addition, the CHIPs act and the Inflation Reduction Act will greatly help the US economy in the long run, but plans like this take years before the effects can be felt, but American’s are incapable of thinking long term.

@Inliytened1 The right voted for a man who actively lies to his base and attempted to change the results of the 2020 election because people have to “put food on the table and make sure we have money at the end of the month”, but seem physically incapable of explaining how he’s going to do that more efficiently than the democrats. The lack of self-reflection is astounding.

Even if this video does not convince you of its argument, I think it brings up some interesting ideas and made me change how I think about numbers. While not mentioned in the video, I think the way commas are used should be changed to better reflect the naming convention they use in the video. Specifically, the commas should themselves be represented with an alternative binary notation and be given the names shown at the 59:00 minute mark. Since binary is so long, this convention I’m proposing would help keep track of how large the number is.

@Ero I concede to your and Leo’s point. University is better for a maths education than no university. Your arguments have persuaded me, and going forward I will soften my rhetoric on this topic. Still, professors should put more emphasis on concept to motivate the theory.

@Ero I haven’t learned much past those topics as of now. I am working on a paper for how to share resources more fairly using bachelor’s level mathematics, so I consider myself an amateur mathematician. You may have a point, but I’m not sure what you could teach on a black board that you couldn’t also teach in an online format. I don’t see how lectures on elliptic functions or singular cohomology couldn’t be taught online in the same way that arithmetic and topology are taught online. My guess is that the number of people who teach graduate level mathematics is low, and people who teach online and who specialize in higher level mathematics is incredibly rare, not that these topics couldn’t be taught in an online format in the future. You are correct about proof writing though, as that requires practice and someone to correct the work. My broader point is that math education should focus on what the math is even about first, then discuss the theorems and the proofs. For example, I had a weak understanding of linear algebra in uni. I knew that determinants could calculate the hypervolumes of parallelepipeds and that a matrix is invertibile iff its determinant was not 0, but I had no idea what these disjoint concepts had to do with one another. It wasn’t until 3B1B’s series on linear algebra where it became clear to me how matrices transform vector spaces in a way which is “linear”, and how the determinant tells you how much the matrix transforms the hypervolumes of unit hypercubes. So when the determinant is 0, that means the hypercubes are flattened, meaning the linear transformation is not one-to-one, meaning the matrix is not invertible. It would have taken my professor half an hour at most to explain this and it would have made every subsequent theorem and proof related to determinants make more sense. There is one class I took which I don’t consider a waste of time and money. It was a topology course I took in graduate school which utilized the Moore method. Essentially, the professor gave no lectures, we were not allowed to read any text books, or use online sources. All we had was a pamphlet which started with two axioms for topology, some definitions based on those axioms, and then statements based on those axioms and definitions. We were not told which statements were true and which were false, and it was up to us to either prove or disprove the statement. As the pamphlet progressed, more definitions and statements were added which were based on the axioms, definitions, and statements prior in the pamphlet. Lectures were essentially students presenting there results to the class. My only gripe with this class is that I would teach it backwards. It started with very abstract concepts and from them it built more tangible structures. For example, it started with topologies, then discussed the separation axioms 1-4, then it went to metric spaces, and finally got to the real number line. I would start with the more intuitive structures, and then chisel away at them until all that remained were the axioms of topology, as this is closer to how topology was created in the first place. This class encouraged critical, independent, and creative thinking, as well as strengthened my proof writing more than any lecture-homework-study-test-repeat style class, and its not even close. If more math classes were like this, I wouldn’t consider bachelor level’s mathematics courses a waste of time and money.

Eh the mathematics education at university was lack luster. It was mostly brute memorization of theorems and proofs with little emphasis on conceptual understanding or intuition for the math being taught. It became more about computation than knowledge. For example, I received a better understanding of linear algebra from 3Blue1Brown’s youtube series on the subject than I did from 2 quarters of uni. I learned more in uni for sure, but I was able to understand the core concepts of linear algebra much better from 3B1B’s free youtube series, which really shouldn’t be the case given how much time and money university classes cost in the US. As another example, I took multiple classes in group theory without understanding that group theory is the mathematical study of symmetry. Group theory felt like this very abstract, rudimentary version of algebra with no solid basis in reality. Had it been explained to me that everything in the class was related to the intuitive idea of symmetry, I would have been more motivated to learn the subject and the concepts would have made more sense since they would be grounded in something I understood rather than it feeling like I’m using logic and algebra to manipulate symbols to acquire some desired, yet ultimately meaningless, result. The modern education system just feels outdated. Before the internet, smartphones, and AI/LLM, it made sense to memorize an enormous amount of facts related to your field of specialization. But in this day and age where anyone can look up anything from anywhere at anytime, memorizing specific facts seems less important. Instead, education should put more emphasis on concepts and intuition as these things are easier to remember and do more to change how one thinks about the world. If your goal is to get a good job which requires a math education, then by all means, go to uni. However, if you want to learn math for math’s sake, there are cheaper, faster, and better resources online and great text books as well. Just my opinion, of course.

@OBEler I wouldn’t use the content of the video personally. I haven’t watched one of Leo’s videos in years, as I already have a pretty good idea of what his thoughts are and his videos are drawn out (always needed x2 speed while watching an Actualized.org video). However, his youtube audience is larger than his audience on this website. My main point is if he endorses the use of AI for higher purposes beyond images of unique pizza toppings, it could inspire others. The use of AI may be obvious to you and me, but a video on how to utilize it properly could be an incredibly important topic for a younger person just starting on their actualization journey.

Existence.

You claim to be one of the few people on this planet who understands time thoroughly so I’d be interested in a metaphysical breakdown into the nature of time (and perhaps it’s relationship to space). My guess is that it would be too advance a video for you to put it on your Youtube channel and it would instead be something you would put in one of your paid programs, but I’d still like to see it on your YT channel if possible.

Perhaps you are still in the process of teaching yourself how fo best utilize AI, but I think a video demonstrating how you use AI as a learning tool could be valuable. To you these things may be obvious, but for many of your students these tips could get help us started with using AI for inquiry rather than just shallow conversation and fun artwork. Probably not as interesting as some of your other videos, but it may be useful. Things like specific techniques and types of questions that people should be asking would be appreciated.