Should I stick with Bitcoin or create my own crypto?

I've got this project in the works, it's a Rust-based AI home assistant that runs on its own. You can check out the details here: Serious manifesto: [link] A bit of a funny take, George Carlin style: [link] Dev branch: [link] If you peek at the manifesto, phase II talks about having a decentralized compute network with a privacy-oriented cryptocurrency. I was initially leaning towards making my own crypto. Monero is solid, but I really can't stand how it's set up with all those daemons wallet, server node, client, you name it. And there's no way I'm touching Ethereum, I'm not a fan of that protocol at all. But I've been working with Bitcoin-related stuff since 2010, so I know the protocol pretty well. I even built my own node once. I might not be up to date with all the latest BIPs, but the core protocol hasn't changed much. Now I'm thinking that instead of going through the hassle of creating my own currency, maybe just using Bitcoin would save me a bunch of time and effort. What do you guys think?

You can't enter a Bitcoin forum and expect people to choose something else over Bitcoin. Obviously, Bitcoin of course. Copying Bitcoin algorithm inside another cryptocurrency won't recreate the same potential as Bitcoin. Those who tried to make replica of Bitcoin including forked coins never succeed, instead of wasting the money, buy and hold Bitcoin.

You won't succeed of making another coin. Bitcoin has already been a successful project since 2009. If you are going to recreate the same thing, no one would support you because there are lots of coins out there with similar features like Bitcoin or Monero. So go with Bitcoin developing. If you are going to help develop Bitcoin, why not join their GitHub as a contributor? Check the link below - https://github.com/bitcoin/bitcoin/graphs/contributors I just want to share these tools and guides below it might be helpful for developing. - https://github.com/igorbarinov/awesome-bitcoin

not true theres tons of new projects out there this year alone that are doing very well this is just the beginning of the industry

I would be surprised if anyone who used BTC for more than 5 years didn't have an idea for designing their own crypto. There have been so many movements for BTC redesign. "BTC needs bigger blocks to scale" "BTC needs better quantum computer encryption" "BTC needs more transactions per second to compete with credit cards" Having witnessed first hand BTC's success, how could they not want to try to build a better cryptographic mouse trap.

Based on the things you said in the description of your project, I'd say Bitcoin is not suitable to be used by it in any form. It's going to benefit you the most if you develop your own completely independent network so that you can have control over its protocol and restrictions that suits your specific project needs.

You definitely need to create your own network protocol. You can't just request arbitrary Bitcoin node to perform specific computation just for you. There are some project claim to create decentralized AI. I expect most of them are either buzzword, centralized or mere theory, but you might want to research about those project.

Technically you can, the question is rather "should you", or even "do you have enough coins to pay for it". Because it is possible to use binary gates, by pushing OP_TRUE or OP_FALSE on the stack, and a NAND gate can be done with just "OP_ADD OP_2 OP_EQUAL OP_NOT". Of course, there are more complex opcodes, which can be used for many computations, to not wire everything with single bits, but in general, if the size of the Script is infinite, then it can perform arbitrary computations (and when a given Script is too long, then it can be splitted between many transactions, and the chain of scripts can be protected by pre-signed transactions with a shared N-of-N multisig Schnorr signature, which would take 64 bytes). Also, pegging in any decentralized sidechain to the Bitcoin network is now possible, by using Proof of Work on top of DER signatures. Then, the merkle tree of the next network state can be simply turned into private key, and it can be used by any sidechain miner to grind it, and claim the reward, directly in BTC. As long as k-values used in signatures are equal to 1/2, publishing the signature will reveal the private key behind it. But as long as all honest sidechain miners collectively have more power, than any attacker, they can safely broadcast their transactions on-chain, and get them confirmed in 10 minutes, if for example double-spending it will require grinding it for hours or days (they can pick any difficulty they want, to make things tick).

Which one is not true? Having a clone version of Bitcoin with the same feature? Why would I go invest in that new project without a unique feature? Anyone can create a clone version of BTC, but the problem is if you think deeply, why would people use that coin if it works just like BTC? No use case? Have you noticed most of the GPU mining these days? They are dying compared to BTC. Look at some scammers; they use pump.fun to create a useless token to scam anyone.

If i understood you correctly, that means all Bitcoin node perform same computation. In addition, i also think OP's project is about generative AI. So AFAIK further change or addition on need to be done so user (who use OP's project) get generated data by the AI.

Build a bitcoin L2 with bitcoin as the native currency for gas.