3 pointsby CodyGaskin7 months ago1 comment
  • dtagames7 months ago
    This sounds disingenuous and misleading. There is no way LLMs can be cajoled into consistent results. It's all prompts, all the way down.

    While this service may add more prompts around prompts around prompts, that's all it can do. Cursor and other tools already do that with rules files.

    It does not help the overall case for AI tools when we promise things they can't deliver or mislead people about how they work. Using words like "learning" and "memory" here in your advertising are examples of that.

    • CodyGaskin7 months ago
      Thank you for your comment and for raising concerns about the accuracy of our messaging. I’d like to clarify what Memoram actually provides, as it’s fundamentally different from just “prompt engineering” or stacking prompts.

      Memoram is not an LLM or an AI tool itself. Instead, it’s a secure, encrypted database where you can store any information you choose—preferences, facts, notes, or other data you want to persist across AI tools. Think of it as a personal, cross-app “memory vault” that you fully control. When you interact with different AI assistants, you can selectively grant them access to relevant pieces of your stored information, enabling those tools to provide more context-aware and consistent responses, regardless of which LLM or service you’re using

      • dtagames7 months ago
        Thanks for the reply and I can see how that might be helpful. It would definitely improve the pitch to say that directly and leave out the statements that it improves understanding or results.
        • CodyGaskin7 months ago
          Thanks. I will certainly make it more clear.