The Octonaut

the accepted terminology

No, it isn’t. The OSI specifically requires the training data be available or at very least that the source and fee for the data be given so that a user could get the same copy themselves. Because that’s the purpose of something being “open source”. Open source doesn’t just mean free to download and use.

https://opensource.org/ai/open-source-ai-definition

Data Information: Sufficiently detailed information about the data used to train the system so that a skilled person can build a substantially equivalent system. Data Information shall be made available under OSI-approved terms.

In particular, this must include: (1) the complete description of all data used for training, including (if used) of unshareable data, disclosing the provenance of the data, its scope and characteristics, how the data was obtained and selected, the labeling procedures, and data processing and filtering methodologies; (2) a listing of all publicly available training data and where to obtain it; and (3) a listing of all training data obtainable from third parties and where to obtain it, including for fee.

As per their paper, DeepSeek R1 required a very specific training data set because when they tried the same technique with less curated data, they got R"zero’ which basically ran fast and spat out a gibberish salad of English, Chinese and Python.

People are calling DeepSeek open source purely because they called themselves open source, but they seem to just be another free to download, black-box model. The best comparison is to Meta’s LlaMa, which weirdly nobody has decided is going to up-end the tech industry.

In reality “open source” is a terrible terminology for what is a very loose fit when basically trying to say that anyone could recreate or modify the model because they have the exact ‘recipe’.

Again, by this criteria the comments section of a Fox News article is a social media platform. There has to be some form of intent. You could use PasteBin to have a conversation, that wouldn’t make it a messaging application.

No, because Lemmy isn’t social media. It’s a link aggregator.

Social media requires you to know who the other people are, or at least that the identity and personality of the other people posting matters to what you consume. Apart from one or two attention-seeking exceptions, I almost never notice who posted something.

In fact, Lemmy being a Reddit clone, you may remember Reddit stirring controversy for years as they did try to become social media - adding avatars, followers functions, chat groups, etc.; none of which really suit the platform or its audience. Perhaps as the audience has changed they’ve gotten what they wanted.

If “social media” is just the ability to comment anonymously on Internet content and argue with strangers, then the guest book on my Geocities soccer page was social media.

It would explain the Nazi salute