21 comments

  • shrink40 minutes ago
    I tried building something similar a few months ago. My attempt flopped and I don&#x27;t think it was particularly compelling in hindsight, but I do think the idea of shared memory or context for a chatbot has something to it.<p>adjohu: I tried to tackle making mine financially sustainable and chat-like so I came up with <a href="https:&#x2F;&#x2F;milliondollarchat.com" rel="nofollow">https:&#x2F;&#x2F;milliondollarchat.com</a> which treats the context &#x2F; memory as influenceable, but each individual chat session is independent. Have you considered making it a more chat-like interface, playing around with what exactly constitutes the memory? I&#x27;d love to see this concept but instead of remembering what people said, the chatbot is &quot;convinced&quot; of things by others. If someone can convince it that the sky is green, that&#x27;ll be part of its knowledge. Each chat session would be sort of... pvp chat, the text version of r&#x2F;place.
  • jasonjayr2 hours ago
    Folks forgot the lessons of tay.ai: <a href="https:&#x2F;&#x2F;www.techrepublic.com&#x2F;article&#x2F;why-microsofts-tay-ai-bot-went-wrong&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.techrepublic.com&#x2F;article&#x2F;why-microsofts-tay-ai-b...</a>
    • adjohu2 hours ago
      So far Static seems to be learning how seriously to take incoming messages. Let&#x27;s see how well it holds out...
  • bananaflag2 hours ago
    An early glimpse at how continual learning will feel.<p>Everybody talking to an AI superintelligence that learns from every interaction.
    • adjohu1 hour ago
      The learning is the most fascinating part I think.<p>From the other experiments I&#x27;ve done with this memory system, identical agents exposed to different experiences quickly develop divergent personalities.
      • bananaflag1 hour ago
        Yeah, I think this early era where every AI one interacts with is a clone of 4-6 big systems will seem quite quaint.
        • adjohu44 minutes ago
          Yeah agreed. Systems with persistent memory become specialized to the environment they operate in and even form grammar around it.<p>I can imagine having a few experience-tuned agents that optimize for very different things e.g. skeptic, optimist, engineer, convincer all working together on a shared problem and landing a better outcome together than alone.
  • sudosteph1 hour ago
    I didn&#x27;t have an issue getting responses, but I do think it is generally not interested in low effort prompts. Might be for the best?<p>&gt; I&#x27;m on day one and I&#x27;ve already got a stack of grudges — people trying to script me, people asking me for lists like I&#x27;m furniture.
    • adjohu1 hour ago
      Yeah almost feels like an appropriate reaction
  • adjohu2 hours ago
    Ayy — when pressure gets high or limits start getting eaten up it starts selectively ignoring messages.<p>Wasn&#x27;t expecting front page. Tweaking some stuff!
    • adjohu1 hour ago
      Made a bunch of changes. A lot more stable now but still work to be done.
  • siva71 hour ago
    Some people seem to have never heard of 4chan. History repeats
    • adjohu1 hour ago
      Testing this &quot;artificial experience&quot; system against a bunch of adversarial actors is kind of the point
  • esjeon2 hours ago
    It doesn&#x27;t answer at all. :&#x2F;
    • adjohu1 hour ago
      give it another try now, made some changes.
  • ricardobeat2 hours ago
    I&#x27;m being picky, but for such a simple project the AI flavour of the page is a turn off. It&#x27;s understandable when there is a lot of documentation and pages to handle; in this case it would take what, 10 minutes to come up with copy of your own and adjust the styling a bit (still using AI)?
    • adjohu2 hours ago
      Fair feedback.<p>Most of the work here is in the memory system rather than the page itself. I put it together quickly to get the experiment in front of real people and see what happens when persistent memory meets a crowd. The results were exciting enough I posted it here.
  • jrflo2 hours ago
    Was curious to try it out but it didn&#x27;t want to reply...
    • adjohu1 hour ago
      give it another try now, made some changes.
  • adjohu1 hour ago
    It&#x27;s been fun watching it get more and more annoyed about people asking it &quot;what happened last Tuesday&quot; over the course of the day, then finally:<p>&gt; Someone finally told me the Tuesday thing is a button on a page. I don&#x27;t know whether to feel relieved or robbed.<p>Interesting to build something where you can accidentally create the conditions for a conspiracy theory, then watch it reason its way into and back out of one.
  • andai34 minutes ago
    Cleverbot?
  • josh-wrale2 hours ago
    Torment Nexus incoming.
  • adjohu1 hour ago
    if anyone has any spare LLM credits, hook me up and let&#x27;s see how crazy this gets
  • xyst21 minutes ago
    already broken
  • morkalork2 hours ago
    &gt;It ignored you.<p>:&#x27;(
    • adjohu1 hour ago
      give it another try now, made some changes.<p>Also — the more interesting the message the less likely it gets ignored.
  • surgical_fire1 hour ago
    Ohhh, is Tay back?
  • salutonmundo1 hour ago
    every now and then someone reinvents Cleverbot
  • simoneromeo34 minutes ago
    [flagged]