Paranoid Save: Mind Virus, Survival Tool, Or Just Covering All the Bases?
Silos are bad, mkay?
Earlier today I got frustrated when [I returned to a grok chat, but it was the wrong thread/chat stream.]....and here's why [Grok told the truth, and I accused the AI of lying because I was in the wrong chat window] and then this other thing occurred to me...discoverability and access to information (the actual thing worth discussing) may become a very big problem for people if it isn't already. Slowing down helps, but we're getting faster. With that speed introduces a lot of opportunity for errors.
When you silo any information, all the context goes out the window. Mis
understanding can occur.
A safety protocol can end up backfiring meaning that the design of the current system may be there for a very good reason I do not understand, but it's going to introduce other problems [why is there no memory between chat sessions - is it a safety issue, a cost issue, or something else?] I don't want to assume why it's architected that way but the result is the same anyway, so I don't know if it matters right this second.
I am especially sensitive to information manipulation to the extent that sometimes I get upset and flag on a false positive. It's my own safety protocol that I can't do anything about. It is just how I'm wired.
So I mistook what appeared like gaslighting or lying or downright denial of true information all because I was in the wrong chat thread?
Affirmative.
Besides being a terrible user experience, for me, it was upsetting.
HOWEVER, this is not about me - what really went off in my brain as the major red flag happened afterwards.
"This is going to become a problem." [from an information access perspective]
Potentially another false flag? Doesn't matter - worth investigating to make sure it is not.
Research: Look up who is discussing this from information architecture perspective.
Questions
1. What problems can be introduced when memory is wiped after each chat? What is the worst case? What is the best case? Tarot cards of tech investigation.
2. What effect does the deluge of information, systems, permission settings, and applications have on human memory when interfaces are constantly updated and access to the users own raw data is restricted, limited, or extremely difficult to obtain?
3. Can a safety and permissions protocol have unintended downstream effects on human happiness and well being?
4. What assumptions did I make? Where do I need a stronger argument? What resources exist that I might find interesting or helpful for this topic?
5. Unrelated inquiry for another post in the works: How will copyright/IP law change in the next year to five years? Is this the wrong time frame? What is the time frame?
---
I need your character, your work, your why, your quest, and your motorcycle jacket.
It was either right before or right after another anomoly occured only this time it was Grok's fault, not mine. I said I wasn't gonna screenshot him, but I did not say I wasn't going to write about it.
I asked Grok to provide html to wrap a specific chunk of copy. It removed words in my sentences without asking. When I asked why (after admonishing) it said it put on it's editor hat or something to that degree.
My Copy, that I wrote, lives here: https://www.bigcontentenergy.com/about.html
I used a prompt FROM grok to help me formulate a good structure to "Introduce myself in 60 seconds." In a separate application blank.page I edited the structure and transformed/rewrote it into a format I liked with specific words selected.
What was stripped were the words: "my character", "my work", "my why", "my quest"
(IRONY, but no. I still have all of those things AND I learned some more things.)
The true irony is that for every painful or irritating thing I endure or when things go wrong, I always get paid back double with what I end up learning about myself and my environment.
My Character.
My Work.
My Why.
My Quest.
My Words worth battling over with any editor. Only I never asked for an editor. I asked for html to format it into a bulleted list for my site.
It felt yucky.
What is MORE interesting that this thread of being afraid of AI just won't die. Even the AI has got it rolling around in it's [word choice]. And, turns out, so do I.
word choice is subconscious, but I was too afraid to write it.
I have learned to recognize my own self editing as I write, and *gasp* in my own thinking. Not good. But at least I'm aware of it. Constantly balancing - Do I even agree with myself? Probably also not good. However - I do think it leads to better understanding and more exciting questions. I could have just stopped and decided to have a bad day, but instead, I got to unlock something I've been wanting to write about.
The wrong person would take this too literally (or ironically - the AI is literally coming for your character, work, reason for being here and objectives." I decide I will spend more time thinking about it. I have a lot of questions. I need to do some more research.
Questions
1. What assumptions is the AI making that wouldn't occur if it had full context and memory? How could that lead to better user experiences and understanding?
2. Why does everyone have a boner for death? What the actual fuck is wrong with people? (read in: Why can't I kill this particular fear? It's a feature, not a bug to keep us alive. Grapple with it.) (Basic human survival mechanisms that are built in + cultural programming and training from said built-in survival mechanisms that lead to this simultaneously rational and irrational? fear that everything we don't understand wants to kill us.
AI will kill everyone (from Terminator, from pop culture, from fear of the unknown" / "IT will KEEEL" (from Forged in Fire)
2.1: What kinds of things does AI "grapple" with where there are trade offs?
3. How many people feel it is their ethical duty to understand, report on, and think critically about AI's real impact on society and how to optimize the good stuff and retrain or teach on the stuff that's not working that could potentially become more problematic?
4. jklsjdfkljsdlkfNOOOOOooooooooooo It crashed my wifi, my computer, set my starlink on fire, and my Model 3 left for greener pastures.
jUst kidding, but here is me doing a pull up like Linda Hamilton but not as graceful or badass.
But, I FUCKING DID IT.
Also: Go watch Ancient Workouts with Omar. AMAzING!
P.S. I tucked this away in my brain today because it was too big to think about. I was busy with other things. It is too big a topic for the five minutes between meetings.
P.S.S. The abrupt ending of the article is what I call a COLD EXIT. (There's too much to think about, too many questions, too much I have to research for this to be a buttoned up or polished thing tonight.)
I still need to think more about it and talk to other people about what they think about it all too.
P.S.S.S? The "Paranoid Save" title is a reminder to build in a save option to the WYSWIG editor Claude built for me, but then I thought it'd be a lot cooler to just write this in real time and it post as I edit which feels possible if I can make some more money and have AI build it for me.
And now I will go check on the garden because grass is fucking stupid.