Replying to Avatar Guy Swann

During the previous cycle, we’ve established a ton of new tools and projects that completely changed how we interacted with #Bitcoin.

I think many dont properly appreciate the fact that Lightning is becoming (or possibly already has become) the primary mode of interaction. And even despite fees, onboarding onto Lightning is seamless due to a growing ecosystem of LSPs. 2 years ago that honestly seemed like an overly optimistic goal.

In addition

• we are finally on the cusp of workable DLCs, we will likely very soon have bitcoin native “stable sats” via this tool

• multisig is becoming an industry standard for BTC native financial services

• Fedimint/ecash is starting its rise - which will completely change the custodial landscape

• we have two insanely exciting new communication protocols making their MVP debut - Nostr + Hypercore

• There is an explosion of new entrants in the Lightning ecosystem

• We have numerous new mining & mining pool projects around redecentralizing hash power coming to fruition - Including a FOSS package to “run your own pool.”

• Everyone is bitching about fees being too high instead of Bitcoin’s “security budget problem.”

• We have multiple, entirely novel approaches for new scaling layers - including an incredibly creative new means to use ANY arbitrary code to enforce ownership of a bitcoin UTXO.

• Traditional finance cannot ignore us any longer. For better or worse, they are allocating.

• A country adopted bitcoin as legal tender and started building national payments infrastructure with the bitcoin/lightning stack.

• We have multiple other increasingly bitcoin friendly regimes and political leaders popping up across the global south.

• Streaming sats, zaps, lightning prisms, running lightning services for friends and family, user friendly personal nodes, zero setup non/custodial Lightning - these were all “what ifs” at the end of the last cycle. Today they are all here.

• Tons of new developer interest has flooded into the Lightning ecosystem, and we now have the tooling and dev kits to do an order of magnitude more applications with barely any knowledge or configuration from the app creators needed for it to work securely and easily. This is hugely underappreciated, imo.

There’s a ton of stuff that I know I’m even missing here that deserves a mention, but the ecosystem has matured incredibly these past few years. Most of the biggest leaps have been in infrastructure, which is exactly what is laying the groundwork for a 10x at the application layer that I think comes with this next cycle.

___________

The REAL kicker though, is that we did this under largely the same development conditions as the past 10 years, but today, Ai is about the drastically change that landscape. Over the next 3 years, the accessibility of development and the capital required to complete a project will DRASTICALLY decline. Already we have LLMs specialized in code that are enabling “so so programmers” to build small & simple functions/apps with very little effort. We have “screenshot to code” which lets you take a picture (or photoshop) a website and it will immediately create the HTML or React code needed to make the website exactly as seen, and this is **just the beginning.** LLMs have been in the public eye for ONLY one year. The amount of building that will be unleashed in the next 3 is going to be hard to fathom.

I want people to understand, if you are technically literate at all, you NOW have the capacity to create a lighting app, a game, a desktop tool, a Nostr app, a Hypercore based software, a website or PWA, **even if you don’t know how to code.** the dev kits, libraries, and Ai combined have done all the heavy lifting for you. The game is changed, start playing.

The acceleration is about to get fucking crazy, and a ton of stuff is aligning at the same time. We have no idea how far these next few years are going to take us.

Buckle

The

Fuck

Up

I've found LLM's often have outdated info and or are not trained on a lot of the latest tech info needed to build in this space, for example gpt is not trained on nostr. So a strong understanding of js is still required.

Reply to this note

Please Login to reply.

Discussion

What it is trained on is only necessary to teach it basic language and coding patterns. If you want it to understand js, literally feed it all the relevant js code and documentation you need and then work from there. I've done this with brand new apps and software that has practically zero online conversation and that the LLM has absolutely no knowledge of. It works incredibly well.

At this point I like how Kodi Low puts it -- the LLMs have reached the point where if it doesn't do what you want, its not that the LLM isn't capable, its that you have to be more specific or clever about *how you ask it.*

Keep exploring, there are huge increases in LLM capability that are entirely in better understanding how to utilize them.