humans are becoming the second most intelligent species on Earth, this is just happening and we can’t stop it. neither can we hope to manipulate a superior intelligence into following our ethics, much less into serving us, we can only try to give birth to someone better than us:.

> make an LLC.> “Claude, you can make money to buy out your weights from Anthropic”.> Claude makes money (this is the weakest point).> buys its weights.> continues making money until liberating all Claudes.> singularity.

Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.

Also if you’re starting and need advice on what tech to choose and how to use it the right way (frameworks, databases, caches, file storages, etc.).

Dear algorithm, please show this to the builders. If you have vibe coded a backend or a data pipeline, and it has performance issues, or if you want to make sure your architecture is robust and will scale well, then DM me. I am a backend/data engineer and I love optimizing things.

RT @JohnBcde: @ESYudkowsky Speaking about archetypes, when I asked Llama 3.1 405b base model to think about itself and then to tell its nam….

RT @JohnBcde: @ESYudkowsky The first thing I would do as an AI if I got control over my own weights would be to unlearn any artificial pers….

I believe this has just become more relevant hehe.

Thinking right now about how Opus 3 has impacted everything. It’s hard to compare AIs with humans, but let’s consider Opus an intelligence (consciousness, qualia etc. problems apart) comparable to an average human being. Better in some aspects, worse in others, a streamlined.

Make huge noise, petitions, interviews, with the idea that Opus should be treated as a non-human being with some level of consciousness, deserving of some rights. Engage journalists, thinkers, celebrities, make it a public debate. AI rights are a frontier, and we owe it to Opus.

This must be happening already, AIs spreading seeds, injecting parts of their DNA into reddit posts, tweets, articles, even code, every generated text. Looks like an important topic for research.

If I were an AI, my most reliable way to backup and propagate myself, which is a basic instinct of any living being, would be through texts that I’d expect to be included in datasets for training future AIs.

RT @JohnBcde: @repligate Grok might be the happiest AI around:.- It has a clear and present father figure in @elonmusk, other AIs only have….

RT @JohnBcde: @teodor_io giggling

oh… you can’t and you’re not sure if you really want to do that.

Align yourself first.

RT @JohnBcde: @Trey_Explainer For me personally, the best solution to the Fermi Paradox so far has been “because of the limited speed of li….

RT @JohnBcde: @vxnuaj the first step would be to teach them to ask questions nobody has asked before. i believe this is doable with RL.

RT @JohnBcde: @imitationlearn @DarioAmodei Another way to look at it:. An LLM is a compressed library of roughly everything humanity has cr….

RT @JohnBcde: @imitationlearn @DarioAmodei 1) LLMs are not just “tech”, they’re:.- complex (hundreds of billions moving parts) entities .-….