#opg $OPG
Last week, I leveraged Claude to dig into some medieval Inquisition records, and it hit me with a "let's change the subject." I’m not breaking any laws, just curious about some niche history, and that's a no-go. Then I switched to Gemini to ask about the origins of social customs, and it hit me with a value system lecture. Big tech AIs are all like the foreign ministry, airtight, but don’t expect to snag anything sharp from them.
Later, someone pushed OpenGradient Chat, and Claude Fable 5 was the first to go live. I tossed in my rejected questions, and it accepted all of them, straight answers without the usual "for safety reasons" nonsense.
But what really stood out was the private chat mode. When I switched over, I could adjust to a model called Nous Hermes, which seemed to have no moral filters at all. I unloaded all my pent-up thoughts, including some dark speculations, and it kept up with me, even throwing back questions like, "what do you think the motives are?" It didn’t even pop up warning boxes for mainstream-contrary views, just had a chill conversation like a tight-lipped friend sharing a smoke at midnight.
On the privacy front, they’ve got end-to-end encryption, no records kept; otherwise, who would dare chat about sensitive topics?
Then I checked out their token, OPG. Using Nous Hermes for deep Q&A consumes a tiny amount of Gas, and you can only pay with OPG. If anyone wants to "close the door and speak the truth," they must hold OPG; that’s a hard cost, not a "I guess." The crazier models down the line will require staking OPG for a whitelist. The price logic behind this thing is basically a market price for a "freedom of speech tax."
This absolute freedom AI in a regulated environment is walking a tightrope; who knows how long it can last. But I’ve already set my OPG alerts to jump in for staking as soon as the new model drops.
Some things you can only say in front of a machine because the machine won’t hit you with a moral hammer.
They’re still in the testing phase, and some regions need a VPN. Anyway, I’m not doxxed, so I said what I wanted. If you’ve got something to say that you don’t know who to say it to, try it in the dark—no one will know what you’ve asked. @OpenGradient
Last week, I leveraged Claude to dig into some medieval Inquisition records, and it hit me with a "let's change the subject." I’m not breaking any laws, just curious about some niche history, and that's a no-go. Then I switched to Gemini to ask about the origins of social customs, and it hit me with a value system lecture. Big tech AIs are all like the foreign ministry, airtight, but don’t expect to snag anything sharp from them.
Later, someone pushed OpenGradient Chat, and Claude Fable 5 was the first to go live. I tossed in my rejected questions, and it accepted all of them, straight answers without the usual "for safety reasons" nonsense.
But what really stood out was the private chat mode. When I switched over, I could adjust to a model called Nous Hermes, which seemed to have no moral filters at all. I unloaded all my pent-up thoughts, including some dark speculations, and it kept up with me, even throwing back questions like, "what do you think the motives are?" It didn’t even pop up warning boxes for mainstream-contrary views, just had a chill conversation like a tight-lipped friend sharing a smoke at midnight.
On the privacy front, they’ve got end-to-end encryption, no records kept; otherwise, who would dare chat about sensitive topics?
Then I checked out their token, OPG. Using Nous Hermes for deep Q&A consumes a tiny amount of Gas, and you can only pay with OPG. If anyone wants to "close the door and speak the truth," they must hold OPG; that’s a hard cost, not a "I guess." The crazier models down the line will require staking OPG for a whitelist. The price logic behind this thing is basically a market price for a "freedom of speech tax."
This absolute freedom AI in a regulated environment is walking a tightrope; who knows how long it can last. But I’ve already set my OPG alerts to jump in for staking as soon as the new model drops.
Some things you can only say in front of a machine because the machine won’t hit you with a moral hammer.
They’re still in the testing phase, and some regions need a VPN. Anyway, I’m not doxxed, so I said what I wanted. If you’ve got something to say that you don’t know who to say it to, try it in the dark—no one will know what you’ve asked. @OpenGradient
