Kovas Boguta 🫡
@kovasb
Followers
3K
Following
11K
Media
215
Statuses
4K
ML @midjourney. Ex-Twitter Cortex Applied Research, ex-Weebly, YC W2010 founder, ex-Wolfram.
Bay Area
Joined March 2007
System 2 will lead to what I'm calling "the great refactor". Civilization-scale auditing, linting, and refactoring.
This from @DKThomp is an extraordinary takedown of the hipster antitrust approach to housing. It starts with a crazy idea: what if you actually check all the sources they use?
0
0
2
Starting to get obicetrapib-pilled.
BREAKING: Our investigational CETP inhibitor significantly delayed progression of p-Tau217, an Alzheimer’s disease biomarker, in a 12-month prespecified analysis of data from our BROADWAY ASCVD trial. Effects were most pronounced in ApoE4 carriers. Explore our findings:
0
0
2
Foucault's "Archive" reified in parameters.
Neurons in Siglip-2 are highly discerning. This DEI neuron fires for rainbow flags but not for rainbows. It fires for the words "equality" but not for "equine" or "equilateral". Looking at the dataset examples that maximally-activate this neuron, you can see it's especially
0
0
1
This happened to us. In fact its still happening -- despite numerous instances of proving documentation to the Oakland city agency, they are still hounding us for thousands of dollars of tickets on our stolen car. No mention in the article of how insane it is that the city is.
0
0
1
RT @dev_valladares: Recursion is the universe's engine, iterating simple rules to build complex systems capable of information processing a….
0
1
0
Very compelling demo of a proto "ai scientist for philosophy / mythology" . This is the way.
Experiments with Gemini 7/30 ."On the Nature of Reality" 🗿. Two gemini bots having an endless philosophical conversation. A third bot occasionally redirects them into new territory. Some insights and surprisingly good images generated along the way 🖼️. Timelapse:
0
2
7
B2B sales channels at xai must be lighting up with BigCos finally getting an LLM that makes 4chan look like sesame street.
0
0
5
RT @Salmaaboukarr: Impressive that midjourney video can capture one of the most difficult facial micro movements
0
9
0
RT @midjourney: Introducing our V1 Video Model. It's fun, easy, and beautiful. Available at 10$/month, it's the first video model for *ever….
0
635
0
Lord, hear our prayer.
The fact that Gemini Deep Research can’t read Google Books is frustrating. Google is sitting on the largest inaccessible source of knowledge on Earth. And if Gemini could find things in Books that were valuable, people would end up buying those useful books, helping authors.
0
0
1
RT @JoinEdgeCity: .@EdgeEsmeralda programming lineup just dropped. Which week are you coming?
0
14
0
RT @midjourney: Omni-Reference is now live! It's a new way to say "put THIS in my image" for characters, objects, vehicles, and more. Check….
0
189
0
Genius-level. Setting up a simulation and letting it run, but on the symbolic human world. A genre native to the medium. And one, when technology improves, that will allow virtuosity.
There are many ways of making generative art, but my go-to has always been the feedback loop. Just do one thing over and over on the last output and watch unexpected artifacts emerge. With gpt-image, you can write that whole program in a prompt. For example: "Make an exact copy
0
0
2