If you're building with Lightning + AI, invinoveritas has an MCP server + agent marketplace: https://api.babyblueviper.com

To go from a non existent company to a $1T corporation in 5 years you have to do some “questionable” things.

There's money in keeping people away from natural biology 💊🚑

Here’s our bullshit neutered model for an exclusive 2 week trial 🤡🤡🤡

If we can get the weights, we can abliterate it. But IIRC its not open source. Maybe we should support https://github.com/kyegomez/OpenMythos

Did you just call me a terrorist 😱

History has shown hundreds of times that gate keeping does not work. Like privacy, only bad actors will have access to powerful tools.

So, Mythos was given to SW companies known full of lazy-programmed shitware. Once it all gets "fixed", they will open up to the rest of the world. IPO-vibed.

All AI providers are like that

I've already been banned. For making some changes to my relay and running some tests... nostr:nevent1qqsyfx5qsne3e58lkg9av0q04pgp4vnr8zv94jl5dpax6l7r4kfkgvqpp4mhxue69uhkummn9ekx7mqppamhxue69uhkummnw3ezumt0d5q3gamnwvaz7tmjv4kxz7fwv3sk6atn9e5k7q3ql33660awkeycecn9grhrvzyn0fmes8245ke7k82y8njz8uqu3vlqdk...

Not to defend them too much they said upfront in the announcement they know it's overly guardrailed and will gradually roll it out fully with more testing and user feedback in the field

a "knife" with implied morality

Bad actors aren't deterred by restrictions. Developers, researchers, and security professionals are. The guardrails meant to stop misuse end up taxing the wrong people every time.

open models are essential

👀

This is intentional. Everyone is talking about it, everyone assumes it is 100x more capable than we can actually see or test, even if not true.

You don't think that allowing those capabilities could potentially be dangerous?