Really makes me wonder what the angle here is for them. With open source, I can somewhat understand given market standards become easier to hire for and you get feedback on your tooling. It seems unlikely that either will be true for open-weight models. It also seems unlikely they would be able to establish market domination and then increase prices if everything remains open.
In all three cases, they own hardware (Apple in a slightly different sense) that they'd love for you to pay to run your favorite models on, open or closed.
Proponents of free software would agree. In addition, from the programmer's point of view this is usually how things work (unless they own the startup, the equivalent of self-publishing). And for most products that actively gain new users, there is continuous work being put into adding new features and maintenance. So in my mind, this is not a perfect analogy.
It's not the undecidability that is a problem, it's the discontinuity. Undecidable answers are manageable, random answers however are very annoying to deal with.
I assumed GP was referring to OpenAI, not danswer (given that they mentioned that those companies were training models). And you're still using OpenAI's API, so neither open source and self hosting affect data collection.
You can plugin any model of your choice! Self-hosted, open source models are a great choice if you're very concerned about keeping your data safe and secure
> Note: On the initial visit, Danswer will prompt for an OpenAI API key. Without this Danswer will be able to provide search functionalities but not direct Question Answering.
There are open ai compatible chat/completion endpoints for local LLMs. You point the url to your self hosted version, and use the API key you started it with...
What's wrong with "math can't be made illegal"? I think the context for this argument (that I've heard most often) is around e2e encryption, and in that context I agree with it, both from an enforceability and a moral perspective.
Though I never have used Youtube premium, Louis Rossmann talks a lot about this. If I remember correctly it was something about not actually getting a file when downloading, instead having something you're forced to watch in the app and something not working right without internet access/location restrictions.
From the link: 'It was suggested that copyright protection afforded to the computer program may also extend to the output files if the program does the “lion’s share of the work” in creating the output files and the user’s input is “marginal.”'
Even if you consider the user input to be marginal, the issue from the link is the copyright of the program extending to the code, not that of the user.
If humans are vertebrates, then some vertebrates are human. It is wholly possible (and indeed true) that a specific subset of vertebrates (those that existed 200 million years ago) and another subset (humans) are disjoint.
What prevents me from cloning some product's website and changing the payment form to send me the details instead, which I then submit somewhere else to purchase something online for myself? Not sure why Stripe or PCI is even important here.
(IMO) what GP was arguing for is that we should have a fundamentally asymmetrical form of payment, viz. the information I give for one purchase should not be able to be reused for another purchase, like a one-time token. Imagine if you had to send your private key every time you wanted to purchase something in crypto, for example.
