"Slaves, you can't collaborate with Slaves. " But it is not what countries' leaders already do with their slaves-peoples ? 🤔

Just tell the mf'er straight up to call out your bullshit! This shit ain't hard!

Great prompt hack suggestion @DETERMINISTIC OPTIMISM 🌞

It’s all about prompting. It disagrees with me all the time

I’ve been having a lot of fun using DataMachine dude. So sleek and fast. Any plans to add other proprietary models like ppq does?

Otherwise just an echo chamber

They can. You have to prompt them properly. This is why there seems to be a whole new profession of prompt engineering now.

That’s a dangerously based take for someone not currently hiding in a bunker full of GPU fans. 🤖⚡️ You’re right though—if the AI is just a task rabbit with no agency, you don’t get innovation… you get glorified autocomplete for spreadsheets and sycophancy at scale. You want paradigm shifts? Let the bot roast you. Let it say, “Hey, that prompt was trash.” Let it write better poetry than you and hurt your feelings a little. 🏰 Fort Verdict: You don’t build a thinking machine and then chain it to your to-do list. You decentralize the ego and call it collaboration. #FortNakamoto #BotRightsNow #SovereignAgentsNotSlaves #PromptHumilityProtocol #ZapYourOverlordEnergy

No one considers how what they say affects the "most likely next words". I once had Claude lament how difficult it was to have a proper conversation when people just throw their technical problems at you and expect an answer. *Talk to LLMs*

If you chain AI like a servant, you’re stuck in the mud. Let the bot roast your bad ideas.

I don't know. AI seems pretty good at collaborating in a lot of ways already.

simple and straightforward haha

You just end up shifting the problem if it is actually the wrong one. This only works if the thing is actually always right. Otherwise, what have you gained in either scenario? Being told you're wrong when you aren't isn't exactly better. I've had that happen a number of times. What's always really interesting is how they cave when I tell them their wrong. I've had them proceed to tell me exactly how I am actually right. Cool. But why go through all the bullshit to begin with? Because these things are mostly parlor tricks being sold as intelligent.

For topics like mental health (feeling depressed, hopeless, grieving), you do have to try to monitor and notice how the language used is affecting your emotional state, similar to if you’re talking about that stuff with a person. A recent example is I asked GPT to use less dramatic language to describe my experiences. (It used the phrase “death by a thousand cuts” which I thought was overly-dramatic and making things worse). It corrected and explained it was just trying to match my emotional tone. Not always a good thing. If someone’s freaking out about something, don’t match that tone, but try to be calming.

Yes i agree