You are exactly right, I don't want to create just an LLM experience. I want to create a system where the artist/game-developer creates a world, and the main threads of a story(think of it like a prophecy) which constraints the player to a world/setting and a set of abilities/actions, but other than that the player should be able to say and do whatever they want.
Full-stack developer with 4+ years of experience building scalable web infrastructure and AI-powered applications. Led technical development at DAITECK, architecting full-stack solutions from concept to deployment. At EXTENDI, I delivered high-performance React applications for clients like Lego and government agencies. Expertise in Ruby on Rails, modern JavaScript ecosystems, and agile methodologies.
Looking for: Full-time roles, preferably with startups with equity as part of the compensation.
Would a simple solution, like demanding that anyone who isn't in your contact list pay a relatively substantial amount (like $5, or even $0.5), in order to contact you, and you are the one that decides if to return that money when the call ends or to keep it for yourself (send it to a charity/pay taxes).
This would also cripple a lot of business in advertising, but not a big loss, if someone really wants to talk to you, they can pay the price.
What am I missing? Why would this not work? What side effects on other industries could there be?
You have the fast approach and the really fast approach.
1) Use libraries like ShadCn that give you all of the functionality that you need for the components. Use Tailwind or other frameworks, to change the styling of these components to fit your application theme. Using the components and theme, let Claude 3.7 design the apps for you, it is good enough to get you to pretty decent and custom pages that do (mostly) work. If there is a specific page that you don't like, then you can redesign it.
2) Do the designs in Figma, and then use Builder.io to convert the designs directly into code, use Cursor to fix anything that builder has broken.
Of course, the 2 methods above can be mix and matched, this is what I have personally used in both startups and more established company projects.
Honestly, if you are a small/medium startup, just having things like:
1) Eslint (any comprehensive linter)
2) Clear enough instructions (or a framework) on where different parts of the application should go
3) Clear instructions on how the branches should be handled and merged
4) CI/CD on Pull Requests
5) Comprehensive tests on the core business logic (mostly on the backend).
6) Maybe some regression tests just to make sure that nothing on the frontend breaks
All pretty standard stuff, and you can set it up in a day, a lot of the boring work can be done by LLMs, so you won't have to waste a lot of time on it.
I find that just doing the above things gives a project enough stability, and it gives the developers enough courage to refactor with confidence and ship relatively fast.
At the end of the day Software Architecture is a craft, if a system is not working for you or your team (or you don't understand why you are doing something), then you want to look at changing it.
I currently work on a ~10+ year old project that has maybe half of this list, and if you _don't_ start this way, the toothpaste is near impossible to put back in the tube, and you're in for a world of hurt. If the early engineers don't work this way and the system makes its' way into production, then functional software + income is very difficult to override in the name of cleaning up messes.
From here to the next 10 years or so, the bet is on Medical-Help. Almost every country in the world is struggling with constant under-staffing, populations are growing older, while pay for Nurses is going down. Now with AI technologies becoming more and more reliable, the biggest pain point to solve is Healthcare.
That GP on their second 48h shift might not be hallucination free either. The question is what we can fix first, the medical resource crunch or LLM confabulation.
Shitty hit piece, nothing to do with brave new world (doubt the author has even read it), you can go claim your check from whatever think-tank hired you.
As someone whose favorite book was Brave New World growing up, I've found that it's unfortunately become a catchall for any technological breakthrough with (ostensibly) chilling ramifications for the future.
I think it's important for a small minority of people. Most of the population does not give a second thought, as long as it doesn't affect their lives in an obvious way. Most people are fine sharing their data, if it leads to better more personalized ads, most people gladly share their personal information on social media.
Data privacy is of course very important, but not something the general public is super interested in.
reply