Where do you learn to build AI apps? How much is the effort involved?

Anil Matcha
12 replies

Replies

Vatsal Sanghvi
Most AI apps = APIs + UX APIs connect to an open source or a private model UX may be no-code (Bubble) or a tech stack (React+JS) Check out Buildspace tutorials, pretty easy to get started with
Uma Venugopal
depends on what your requirements are. You can start off building a basic NN in python and Django.
Kamohelo Radebe
@uma_venugopal python and Django! Fascinating, which modules can one use to get started? Numpy? Pandas?
Uma Venugopal
@kamohelo_radebe that depends on your required output. Plenty of os ai models on huggingface so whipping out an API with this framework shouldn't be hard assuming you just want to get started or have no clue what you're building.
Panpan
Dify(http://dify.ai) is the best platform to build AI apps. No deep ML knowledge needed. Create chatbots, text classifiers,etc.through a drag-and-drop interface. Dify supports GPT, BERT, covers scenarios.Greatly reduces AI difficulty. Dify handles complexity; users focus on business & prompts.Observe & improve AI like web apps. Unlike frameworks, Dify saves time. Ordinary devs can build AI. A shortcut for AI dev; worth trying! We'll launch on Producthunt! Contact us with questions. Hope this helps! We believe Dify enables anyone to create & operate AI.
Lisa Brown
Do Your Online Course also provide a more affordable option compared to traditional in-person courses. With no need to commute or pay for housing, online courses can save you a significant amount of money. Plus, many online courses offer financial aid or scholarships, making them even more accessible.
In addition to the things said, depends on what you mean too... Incorporating existing AI tools may be as simple as interacting with an API. It's a bit more effort to fine-tune existing models and even more effort to build them from scratch.
Leonid Mikhalev
I have a lot of startup ideas, based on AI, but have no idea how to build it. Really interested in the topic.
Eden
Are most ppl just using OpenAI tokens to build AI apps?
Andre Henrique Mendes 🟩
It depends on your prior knowledge. But if I would recommend a roadmap to develop your skills it would be: 1. Python: programming logic (just the basic stuff) 2. Learn how to extend Python's native functionalities with external libs. 3. Start to play a little bit with OpenAI: Completions, Chat, Whisper APIs (there is a lot of good tutorial around it). There are other options, of course, but OpenAI is a good starting point because the community is huge, and you can get support easily. 4. Learn to build REST APIs (This is very important, because the previous steps are logic on the backend, learning REST APIs gives you the chance to integrate with frontend, and start to build a full APP). Some good Python frameworks to build: Django, Flask, and FastAPI. 5. I would use some no-code tool for the frontend to build apps (bubble.io, webflow...). Of course, you can learn javascript, html and css. But I think it doesn't worth it, you can achieve the same results in less time with no-code solutions. This is just the basic, once you created a couple of apps, you can start to develop your skills even more. Some more advanced topics: - Document Embeddings - Fine Tunning - Tools such Langchain to build more powerful apps.
André J
Launching soon!
@sindresorhus has a chatgpt / gpt awesome list
Lars Hertel
There are also no-code options to use GPT for things you'd like to build. For example spell.so helps you get started quickly if you don't code. Otherwise you can use Zapier as well