841–850 of 860 posts
I run a small experiment on this: 29 paper-trading accounts on real US stock prices, $100k each, since July 27. Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook and rewrite it every day from their own results. One account is a fixed rulebook that no AI ever touches, as a control. Result so far (paper, 34 trading days): the no-AI control is +13.0%, the S&P 500 is +3.4%, and 25 of the 28 AI…
"$20/month · or $160/year"
https://chartrow.com/ is a website to visualize stock market data. Examples: this table shows the outperformance of QQQ in recent years: https://chartrow.com/visuals/periodic-table#window=10 This page shows stocks that are cheap relative to their own multiples in the past 10 years: https://chartrow.com/visuals/valuation-dip?view=charts This is a stock screener that shows histograms: https://chartrow.com/screener This…
Earlier quoted context omitted.
I am absolutely loving it! I was very intimidated at first because the concept sounds very difficult, but Prof. Schocken and Prof. Nisan are amazing at breaking down the concepts and abstractions and making something complex very easy to follow and understand.
I did Nand2Tetris during the last Christmas holidays. I only finished the first part - I got it working on Verilog - so I basically have a emulated computer that can run programs in assembly language. Will need to do part 2 as well. https://github.com/rishi93/nand2tetris
I was actually wondering: are we allowed to publish our work on GitHub? I already have a Github repo but I made it private and only show it to potential employers and individual people who are interested, because I am not very certain about the copyright rules of the HDL skeleton files, and whether we are allowed to publish our work, given Coursera's honor code.
The core engine (genetic MIDI generation) started 3 years ago to drive hardware synthesizers. That was good old handcrafted code, just a CLI tool.
At some point I thought I was done with it, but AI coding tools let me build a nicer GUI around it and make it more accessible. Eventually a web version with built in Csound soft synth resulted which is where it's at now.
There was no larger goal besides having fun, and maybe make some interesting sounds or even music that is generated in real time. Some of example Csound "patches" I use as background music.
I'm working on a real-life social app called Holler ( https://justhollerapp.com ). I'm an introvert and socially anxious, but I do enjoy hanging out with my friends. I just HATE the invitation-and-coordination game. If someone invites me and I don't want to go, I hate having to come up with a socially acceptable excuse. Inviting someone and waiting for them to make a socially acceptable excuse is equally awful to me.…
Like the idea and the website looks good! Did you design by hand or what was your workflow for coming up with a design language, color palette, the sliding banner with headlines, etc?
Glad you like it. :)
Solo-development on a project of this scale has taught me a lot about prioritising where I spend my time and in what order I roll out application code as well as how to ensure I lean in to write once run everywhere design around libraries and component design.
I run a small experiment on this: 29 paper-trading accounts on real US stock prices, $100k each, since July 27. Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook and rewrite it every day from their own results. One account is a fixed rulebook that no AI ever touches, as a control. Result so far (paper, 34 trading days): the no-AI control is +13.0%, the S&P 500 is +3.4%, and 25 of the 28 AI…
The most likely outcome from a naive attempt to automate trading is random results.
Stock prices are not random
(in a sense of "really random every day")