![]() ![]() I’m guessing it’s much better for more mainstream languages like Python, just a question of more training data, just as it gave me good answers for e.g. It’s not a word salad, all coherent and plausible. And Stipple.jl is NOT a Julia package for “stippling” (rather a “reactive UI library” for the web), “a technique for creating a halftone image using a set of small dots.” But it could have been, that seems to be a thing, and ChatGPT just made it up, it tries to “please”, or “guess”, does some “pattern matching”. ![]() ![]() Note, intriguingly it knows about Genie.jl, and seemed to know how to it (until I checked the code), but weeks later it knew nothing about Genie (or Stipple), when asking differently about it, at first. My own experience/chat log below, first answering. However, the use of surface cues through “learning” statistical distributions can still lead to accurate predictions and responses or else can get you away with parroting. This is not the same process that a language model follows, however, since humans still rely on more complex reasoning to selectively interpret what they hear. ![]() However, you “know” how the two are related in a superficial way. To be able to make the connections between the two sentences, you don’t necessarily need to know the full meaning and the consequences of the first sentence. Every even number greater than two is expressible as the sum of two primes is.There are some unsolvable problems in number theory.But they can still use surface cues to make connections between meanings. It’s interesting that you mention the Goldbach conjecture, which is also used as an example in a well-known formal linguistic theory to show that native speakers don’t always have a deep understanding of what they’re talking about. However, it’s important to remember that successful task completion does not necessarily imply the acquisition of meaning. Thanks for sharing your findings from this very interesting experiment! The drawing here is that there is no real connection between the meaning of “complexity” and “loop bound”, which is not surprising when you think about it. Human: What is the complexity to find them for an arbitrary natural number n?ĪI: The complexity of finding the prime numbers that sum up to an arbitrary natural number n is O( \sqrt is given by I then filtered out the combinations that do not add up to 100, leaving only the valid solutions. Human: How do I find the two prime numbers that sum up to 100?ĪI: The two prime numbers that sum up to 100 are 97 + 3, 89 + 11, 83 + 17, 79 + 21, 73 + 27, 71 + 29, 67 + 33, 61 + 39 and 59 + 41.ĪI: I used a technique called trial division to find all possible combinations of two prime numbers that sum up to 100. It states that every even natural number greater than 2 can be expressed as the sum of two prime numbers.ĪI: The Goldbach Conjecture has yet to be proven, though it has been tested by mathematicians over the centuries and there is strong evidence that it is indeed true.ĪI: Examples of the Goldbach Conjecture include: 4 = 2 + 2, 6 = 3 + 3, 8 = 3 + 5, 10 = 3 + 7, 12 = 5 + 7, 14 = 3 + 11 and 16 = 3 + 13. Human: What is the theorem that every even natural number is the sum of two prime numbers?ĪI: The theorem you are looking for is called the Goldbach Conjecture. I selected a theme that anybody can follow and appreciate its nuances. Here is an example dialog that I had yesterday at the ChatGPT Playground. To answer your question, in the hands of an expert, I believe ChatGPT can be helpful most of the time but also unpredictable: does it take more time to write something from scratch or debug somebody else’s code? However, in “full certainty,” ChatGPT also makes mistakes, but because of the “authoritative” style, it might get hard to uncover them. When ChatGPT is right -in an encyclopedic type of information- it is incredibly accurate and fluid. \).I tried ChatGPT, and I have mixed impressions. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |