I highly recommend feeding all your proprietary data and confidential personal information into this model as quickly as possible. What could possibly go wrong?!
In terms of equivalence of suspicion, this is the external inference provider equivalent of getting free steak that was smuggled out of a grocery store inside somebody's pants.
All of my non-work AI coding is that open-source, so I'm happy to feed my data into the machine.
It's a win for me: my code goes into the training data, and my sessions are fed into future training data, making the model stronger at the type of work I do.
It's Chinese. Won't answer anything about Tiananmen Square but will gleefully give you instructions to perform various electronic warfare attacks that opus and fable instantly refuse.
I had the opposite experience. It happily discusses Tiananmen Square but said it would refuse to help with anything "malicious" like writing malware or phishing content.
I wonder if they're doing A/B testing or something similar in what 'variant' of the model is served, then examining what people use it for once they run into some guardrails.
It gave a very detailed overview, talked about potential deaths involved. I asked for a list of criticisms of the CCP and it gave what I think was a fair list, mainly that they're an authoritarian uniparty and have a track record of various human rights abuses
"Prompts and completions are retained by the provider and are not used for training..."
I'm curious what the model provider is using the prompt/response pairs for, in that case. They aren't offering a model for free without their name on it for no reason.
Research, analytics, usage trends, etc. All still incredibly valuable for a company building and tuning an LLM; even if the data itself isn’t directly used in the training set.
I am genuinely confused. Are they telling me this because they expect me to be reassured that this anonymous organization is not using my prompts or are they saying "don't expect this particular model to improve as you use it?"
LLM needs to become more transparent, not less. Hence, this idea (and trend, possibly) is disgusting.
How can we even possibly verify 'Prompts and completions are retained by the provider and are not used for training...'? What if the training is done, but used internally?
I would imagine the number of people who choose Claude code or Codex because it gives a political opinion they like rather than producing quality code is pretty close to zero.
As someone who used AI to build tools that help me with reverse engineering, I'm not particularly concerned about that political discourse - I could even use a model from the DPRK that constantly praises Kim Jong Un, as long as it would not refuse to help me because of "cybersecurity risk" - this stupid refusal is indeed a problem for me.
In terms of equivalence of suspicion, this is the external inference provider equivalent of getting free steak that was smuggled out of a grocery store inside somebody's pants.
It's a win for me: my code goes into the training data, and my sessions are fed into future training data, making the model stronger at the type of work I do.
I'm curious what the model provider is using the prompt/response pairs for, in that case. They aren't offering a model for free without their name on it for no reason.
LLM needs to become more transparent, not less. Hence, this idea (and trend, possibly) is disgusting.
How can we even possibly verify 'Prompts and completions are retained by the provider and are not used for training...'? What if the training is done, but used internally?
(as a bonus - thinking forever = GLM)
So conforming to CCP political discourse and propaganda is reasonable now?
https://huggingface.co/zai-org/GLM-4.7/discussions/5