Sign In

⚠️ What does open source mean in the age of artificial intelligence?

Haebom
Tap the subscribe button, and you'll get valuable updates delivered to you every day.
Leaving a comment or some emojis would be a huge support. And if you share this as well, that's even better!
Companies are increasingly switching to non-commercial licenses to protect their intellectual property and avoid open source obligations (even GPT-4 is like that).
Even among open source projects centered around large language models, more and more are adopting licenses that only allow non-commercial use.
But this trend misses the point that most of the advantages are focused on academia or hobbyist benefits.
Just saying 'This is a business and an investment, so why not make it public?' isn't enough to cover all the arguments.
For companies, releasing an AI model they've poured resources and effort into for free could hurt their shareholders, so there's no clear reason to make it open source → After all, OpenAI still needs to pay salaries, right?
Going forward, value creation in the AI field is likely to mirror the pattern of innovation, requiring significant capital investments such as putting money into next-generation AI technologies like large generative models and attracting specialized talent.
Major companies will provide the AI backbone and create the ecosystem's foundation, while startups building on top of this still have plenty of room to grow.
But it's thought to be practically impossible for open source projects to outpace big players like AWS, or for the open source community to become a serious rival to OpenAI’s GPT or future models.

In summary

It's true that understanding and interest in open source are necessary in the AI world.
→ Remember, there were lots of fresh attempts when Llama went public.
Companies simply can't help but carefully weigh the impact of their open source licensing choices.
Startups building on tech restricted by non-commercial licenses should be cautious → It might actually be smarter to focus on commercially available models or look to friends who have chosen current game engines (like those who went with Unity or Unreal for their business model).
Because LLMs are expensive by nature, big companies inevitably have a dominant position in the AI space.
From a startup's perspective, it's important to approach future roadmaps and strategies thoughtfully.
Subscribe to 'haebom'
📚 Welcome to Haebom's archives.
---
I post articles related to IT 💻, economy 💰, and humanities 🎭.
If you are curious about my thoughts, perspectives or interests, please subscribe.
haebom@kakao.com
Subscribe