Talk about the two daysโ hot
#Tapeout project.
People have no direct interest hereโI didnโt buy any crystal transistors, and I also didnโt buy $BEM. But when I encounter something I donโt understand, Iโm used to instinctively pua myself: โIf you canโt even understand this, what else can you do?โ
However, Iโm extremely lazy. So I chose a seemingly โresearch-basedโ method that, in reality, is still something I donโt fully understand: hand it over to AI.
This is basically the same approach as how, on social media, you see some piece of knowledgeโignore whether you really get it, just hit favorite/mark it, and youโve already learned the โpoint.โ
After reading the feedback AI gave me, I think I roughly get whatโs going on. Itโs like a LEGO-brick game: you arrange and combine parts to build various โroads.โ If your road can reach the destination accurately while keeping the route the shortest and with the fewest bends, then you get bonus Buffs that help you dig up more BEM tokens, and then sell them for money. Building roads consumes parts, and parts cost money to buy.
So from this angle, itโs still an investmentโminingโgetting ROI or making money. But that doesnโt imply any value judgment. In crypto, even pure โair coinsโ can be traded, and theyโre even easier to trade.
On the other hand, itโs interesting that this building process is a competition. You struggle to come up with the most optimal road, and then a day later someone else builds a road thatโs even better than yoursโyou can hardly not feel compelled to surpass them, and thatโs when the competition really starts.
I asked ChatGPT to try to find optimal roads for a few problems within a limited number of parts, and see if it could get first place on the leaderboard. The 100U/month version of ChatGPT ran over 60% of its usage and still didnโt succeed. Over the whole run, it spent close to 20 hours in total. I manually paused it today. I give up.
Thatโs my Tapeout experience above.