Those experimenting with ChatGPT (evil), should understand that they are using your prompts to train their AI, and that all the information is logged. OpenAI has recently been in the news regarding having let go its ethics and safeguards team in its entirety. Do not be surprised if in the near future we find out that they have been logging all your queries and uses, maybe feeding them directly to authorities. Hell they may already be getting paid by said authorities for data feeds on individuals trying to get the LLM to answer questions related to your grey and black area queries.
Many members here that are ESL and others those who can't write coherent sentences would benefit from asking an LLM to rewrite their posts for clarity. Alternatively you could use a AI writing tool that integrates into your browser like Grammarly. Here is a rewrite of my first paragraph on this post, using a local copy of an LLM called the Mistral 7B model I have running locally in a Docker container on my craptop:
When you use ChatGPT, you're essentially training its artificial intelligence (AI) by providing it with prompts and information. This means that everything you type is being recorded and stored somewhere. Recently, OpenAI, the company behind ChatGPT, made some changes that might raise concerns about your privacy.
What's concerning is that OpenAI used to have a team dedicated to ensuring their AI was developed in an ethical and responsible way. Unfortunately, they got rid of this team entirely! This means there are fewer checks and balances in place to guarantee your data is being handled securely.
As a result, it's possible (or even likely) that your queries, including sensitive topics, might be accessed or shared with authorities without your knowledge or consent. It's like having a private conversation that's actually being monitored by someone else.
This isn't just speculation; there have been instances where AI systems like ChatGPT were used to collect and share personal data without users' awareness. So, it's crucial to be aware of these risks and take steps to protect your privacy when interacting with AI models.
Remember: When you interact with ChatGPT or any other AI system, you're creating a digital footprint that can have real-world consequences!
p.s. how cool would it be to get API access to this forum and feed it to a AI/Deep Learning/LLM to generate a SoSuave chatbot we could ask questions from that would be answered from the decades worth of posts in this forum, exclusively?
gizmodo.com
