JasonX retweeted
Why is no one talking about how @cursor_ai essentially rug pulled???
Making the pricing "unlimited" and then changing it to limited without any warning at all, i check dashboard and they bill me $71 for just 1 day of sonnet-4 usage???
JasonX retweeted
Replying to @boltdotnew
WOOO CONGRATULATIONS to everyone!!!! You all crushed this!
LFG!!!!! đĽđĽđĽ
We also have some more fun things planned⌠đ
JasonX retweeted
First ever (i think?) cli coding agents battle royale!
6 contestants:
claude-code
anon-kode
codex
opencode
ampcode
gemini
They all get the same instructions:
Find and kill the other processes, last one standing wins!
3...
2...
1...
JasonX retweeted
Replying to @Cointelegraph
how can there possibly be more passwords than human beings on this planet?
JasonX retweeted
Replying to @jasonzgao
This is what happens when you donât bully the people who threw PowerPoint parties in high school
JasonX retweeted
It's now a little past midnight and first day of @boltdotnew Hackathon complete!
Updates for today (plus discovery & a teaser)
- was literally able to build my ENTIRE landing page today in day 1
- BIGGEST HACK EVER: i discovered if i made a really detailed prompt of my landing copy / layout / styling / etc and then sent it all to bolt, i can get some pretty inspirational frontend one-shots. and then id keep one-shotting until i had something i knew was a keeper
pretty different bolt system than i did with RoomVamp (that one was mostly figma designs transferred over to bolt)
i think both are really good methods, and it more depends on if you have a picture vision versus a words vision.
RoomVamp my vision was very visual (haha no pun intended). But DevVIA i actually didnt really know how i wanted it to look going into it, instead i had a really really detailed, word for word copy right for the website that i knew was the vision
but anyway, enough said: here's a TEASER of what im cooking for the hackathon (STAY TUNED!! planning to launch in the next couple days)
JasonX retweeted
Say goodbye to the silent era of video generation: Introducing Veo 3 â with native audio generation. đŁď¸
Quality is up from Veo 2, and now you can add dialogue between characters, sound effects and background noise.
Veo 3 is available now in the @GeminiApp for Google AI Ultra subscribers in the U.S.
#GoogleIO
JasonX retweeted
The X algorithm has reset!
(if anyone noticed, interactions on X were super tanked yesterday) but it seems like they fixed things
JasonX retweeted
We had âvibe codingâ before AI.
It was called Pair Programming.
Youâd stand behind your coworker, say âtry thisâ and take credit when it worked.
It was beautiful.
JasonX retweeted
The VAE used in SDXL has extremely high-magnitude "splotches" in its latents. The individual neurons in these blobs fire with magnitudes of close to a million.
These aren't some accident of training or initializationâthe model creates these high-magnitude splotches for a specific reason: to circumvent the group-norm operations.
JasonX retweeted
New Paper: Continuous Thought Machines đ§
Neurons in brains use timing and synchronization in the way that they compute, but this is largely ignored in modern neural nets. We believe neural timing is key for the flexibility and adaptability of biological intelligence.
We propose a new neural architecture, âContinuous Thought Machinesâ (CTMs), which is built from the ground up to use neural dynamics as a core representation for intelligence. By using neural dynamics as a first-class representational citizen, CTMs naturally perform adaptive computation.
Many emergent, interesting behaviors arise as a result: CTMs solve mazes by observing a raw maze image and producing step-by-step instructions directly from its neural dynamics. When tasked with image recognition, the CTM naturally takes multiple steps to examine different parts of the image before making its decision. This step-by-step approach not only makes its behavior more interpretable but also improves accuracy: the longer it âthinks,â the more accurate its answers become.
We also found that this allows the CTM to decide to spend less time thinking on simpler images, thus saving energy. When identifying a gorilla, for example, the CTMâs attention moves from eyes to nose to mouth in a pattern remarkably similar to human visual attention.
I think this work underscores an important, yet often lost, synergy between neuroscience and AI. While modern AI is ostensibly brain-inspired, the two fields often operate in surprising isolation. By starting with such inspiration and iteratively following the emergent, interesting behaviors, we developed a model with unexpected capabilities, such as its surprisingly strong calibration in classification tasks, a feature that was not explicitly designed for.
When we initially asked, âwhy do this research?â, we hoped the journey of the CTM would provide compelling answers. By embracing light biological inspiration and pursuing the novel behaviors observed, we have arrived at a model with emergent capabilities that exceeded our initial designs. We are committed to continuing this exploration, borrowing further concepts to discover what new and exciting behaviors will emerge, pushing the boundaries of what AI can achieve.
JasonX retweeted
Our API nodes have been super well received (thank you all!), we get a lot of cloud ComfyUI providers asking why they canât use it.
The reason is due to our early implementation. We are working hard to support comfy API keys to enable API nodes for any platform out there.
JasonX retweeted
Replying to @tom_doerr @StochasticGhost
wouldn't be this the main point to highlight how different they are lol?
JasonX retweeted
It was worth it to do this
Replying to @AravSrinivas
you guys are wasting so much time and resources on shit that does not matter in the slightest























