Keith
Verified account This account is verified. Learn more@gnukeith
376 Following 22K Followers 18K Posts
Latest posts 5
I went to college to study CS +ethical hacking as they called it, I just wanted to break into systems (I am a Kevin Mitnick fan), you have no idea how many books I owned / read (I didn't read all of them). I was obsessed with learning and absorbing information, I've always loved researching and figuring out better solutions to problems we have today even though sometimes they aren't problems but I feel like they could improve, basically I love doing things from scratch to understand everything from the ground up, I also do believe it's also the high of creating something you fully understand from scratch that I really like.
Anyway, I remember that my professor said that we are being taught to prompt AI's in the future. Now this was back in GPT 3.5 days and you could barely get a sensible Python script out of the model.
Later I had I think I had 1-2 semesters left and I decided to quit and the reason why was very simple. I was extremely bored and I lived outside of the city and didn't want to drive there every morning.
Physics, math, programming, networking were things I absolutely loved (not you networking), the more I learn about networking the more I hate it.
For the rest I felt like they were a waste of time and still to this day do believe they are and I am very glad my parents allowed me to make the decision of being at Brave and Obscura instead of college.
They didn't really understand what was going on until I got relatively popular, people know me in real life????
I share a lot of views with cybersec people and some of my best friends are deeply in the cyberpunk world, some are programmers and some are still in college / university (hello guys!)
It was a dream come true to get the opportunity to work at Brave and I can't believe they kept me.
Same with Obscura, especially early this year when I was starting to have medical issues (still have them now, MRI soon!) so thanks to both companies for supporting me in this time and giving me time to thrive.
I wouldn't call myself a programmer anymore, I'm more of an architect + researcher, I really like creating the backbones for the theoretical.
I love the word theoretical because it can be whatever you want.
Is it theoretically possible to put glass in a water cup? Is it practical? Maybe?
As I find myself becoming increasingly obsessed with AI but not in the sense of, I want it to code for me, but in the sense of discovering more and being able to look at a problem from multiple angles, we are incredibly naive and tunnel vision prone creatures btw. That's probably why we work best in groups.
This is basically a long post of saying, I've enjoyed researching, experimenting and finding practical solutions to make LLMs smaller yet smart. We can achieve what you guys call "SOTA"? Whatever that means on less than 8gb of RAM, however going there requires a lot of time and compute.
Remember to be able to create something small you need a big solution. Look at the ASML machines.
Anyway, I remember that my professor said that we are being taught to prompt AI's in the future. Now this was back in GPT 3.5 days and you could barely get a sensible Python script out of the model.
Later I had I think I had 1-2 semesters left and I decided to quit and the reason why was very simple. I was extremely bored and I lived outside of the city and didn't want to drive there every morning.
Physics, math, programming, networking were things I absolutely loved (not you networking), the more I learn about networking the more I hate it.
For the rest I felt like they were a waste of time and still to this day do believe they are and I am very glad my parents allowed me to make the decision of being at Brave and Obscura instead of college.
They didn't really understand what was going on until I got relatively popular, people know me in real life????
I share a lot of views with cybersec people and some of my best friends are deeply in the cyberpunk world, some are programmers and some are still in college / university (hello guys!)
It was a dream come true to get the opportunity to work at Brave and I can't believe they kept me.
Same with Obscura, especially early this year when I was starting to have medical issues (still have them now, MRI soon!) so thanks to both companies for supporting me in this time and giving me time to thrive.
I wouldn't call myself a programmer anymore, I'm more of an architect + researcher, I really like creating the backbones for the theoretical.
I love the word theoretical because it can be whatever you want.
Is it theoretically possible to put glass in a water cup? Is it practical? Maybe?
As I find myself becoming increasingly obsessed with AI but not in the sense of, I want it to code for me, but in the sense of discovering more and being able to look at a problem from multiple angles, we are incredibly naive and tunnel vision prone creatures btw. That's probably why we work best in groups.
This is basically a long post of saying, I've enjoyed researching, experimenting and finding practical solutions to make LLMs smaller yet smart. We can achieve what you guys call "SOTA"? Whatever that means on less than 8gb of RAM, however going there requires a lot of time and compute.
Remember to be able to create something small you need a big solution. Look at the ASML machines.
So am I safe to assume that this model has an image gen?
New stealth model 🚨View quoted post on X
“Tidepool Alpha Free”, on OpenCode Zen. Let’s see who you are 🔎