Jessica Rose 🤙
3 subscribers
7 photos
Applied Mathematician | Immunologist | Computational Biologist | Molecular Biologist | Biochemist | Musician | Writer |
Download Telegram
Channel created
I just walked to the sea and did my best to intention the winds to die down in California. I cannot express how this much I am feeling this - not sure why. The animals, mainly.

If nothing else, if you read this, set your mind to let the winds the die down and for the fires to burn out.

By the way, if anyone needs a place to stay for free, I have a room. It's a bit far away but it's here.
My friend and brother in arms .@RobertKennedyJr is now the Secretary of Health and Human Services.

The bad guys have no idea what's coming. But we do. And it involves transparency, fairness and justice.

For my part, I am here to make good on my promise to do my very best for the people of America and the world.
I don't believe in the "somebody that I used to know" phenomenon. It's a trick to glamorize loneliness. We know who we know forever in spite of ourselves. Those who are lucky enough to understand this, reap the benefits of true intimacy and transcendent love and friendship.
Good times today with Joseph and his wife and students. :) @Honest_Medicine
I literally look like a 50-year-old child to myself. 😄
On multimodal AI capacities/breakthroughs, self-awareness and stuff:

"Human Knowing: You know a dog through senses (sight, touch), emotions (love for your pet), and reasoning (dogs evolved from wolves). It’s tied to lived experience and a mental model.

AI “Knowing”: The AI’s version is a shadow of that—a data-driven simulation. It’s like a blind artist who’s memorized every description of a dog and can paint one flawlessly but has never seen or felt it."

I disagree with @grok here on its analogy to a blind artist if the artist is human. A blind human artist sees and senses in ways that are in some ways superior to a sighted one - touch, smell, and third eye stiff.

I'm not sure what a better analogy would be yet. I don't think an AI could "human know". Another thing I wonder about is if prompted with the words "Newfoundland dog", would the multimodal AI use the combo LLM and regenerative stuff to make a new image, or would it simply output an existing photo of a Newfoundland dog taken by a human that is online. And if the former, wouldn't the latter be far more efficient and "realistic"?

Can you guess which photo below was AI-generated?