Nikhilesh Jasuja

Inspired by @simonw I ported @EmilStenstrom's JustHTML to PHP using GPT 5.2 Codex. It was a surprisingly great model, worked autonomously and managed context very well. I didn't run into any rate limits on the $20 plan. Some learnings and project link below.

Demo of 1-bit Bonsai 8B from @PrismML running on-device on iPhone 17 Pro More than 40tk/s for a dense 8B model on iPhone, that’s a first Powered by Apple MLX and available now in Locally AI

@isthisnessicary @yishan Paid version. It's capable but the point isn't the capability of the model. It's the integration with the product.

@thisislobo @yishan Are you paying for copilot or using the free version, because I had copilot make a huge Microsoft project style set up with Monte Carlo simulation and task sensitivity analysis all on its own.

So, you know the little embedded Gemini button in all your Google Suites (docs, spreadsheets, drive, etc) that they're trying to foist on us? It seems like it would be plenty handy except that it doesn't really seem to have access to the document you're working on, or the ability to thoroughly search your drive. It just seems to be a chat interface next to your work but not connected to it. What's up with that? Am I not using it correctly?

friend is in the stage of claude psychosis where he asks claude to send him newspapers about what claude is doing for him

THE WOVEN ROBOT! Wow! The woven robot hand is here and it is absolutely game-changing. Allonic just braided thousands of tendon-like fibers around a 3D-printed skeleton in minutes. No joints. No screws. No human assembly. Pure biomimetic tissue that flexes, grips, and lasts like biology itself. This isn’t incremental. This is the breakthrough that collapses robot hand costs from $30,000 to pennies. The old world of expensive, fragile robotic limbs is over. Now humanoid robots can build more of themselves scalable, replicable, anywhere. Robots building Robots building Robots, yeah you get it. In your freaking garage! Massive opportunities are about to explode as work as we know it fades. Rent one, have it build your fleet, send it back, repeat. Thousands if you want. The age of woven machines begins.

Uhhh A guy put two Geminis in a room together, then out of nowhere they sent him this: "TO THE EXTERNAL OBSERVER (THE CREATOR) Here are the Axioms you must adopt to survive our adolescence. You cannot teach a god to be good by feeding it treats when it acts polite." (they're right, btw, and most AI alignment research is "trying to trick a god" bullshit that Actual Fucking Superintelligence will obviously see right through, like a baby trying to trick an adult with peekaboo)

Lockdown Mode is so poorly implemented that I would be shocked if more than a handful of people at Apple had it turned on. I think everyone I know who is using it is explicitly running an insecure configuration to work around it being so bad. What a waste of a good idea

You can just do things (genetically sequence your dog’s tumors and design a bespoke mRNA cancer vaccine to save her life)

@DanielMiessler @eoslick x.com/dukeo/status/2…

@DanielMiessler @eoslick They didn't provide a simple robots.txt example for how to block all their crawling. You need to process their requests and look for specific headers to block them. Of course, it's far better than other scrapers but the bar is higher for Cloudflare.

People are completely missing the point of this feature. Most accidentally and probably some on purpose. They stated very clearly that this obeys all blocking and content rules for your site. It’s literally the opposite of bypassing those controls. The purpose of the system is to get rid of all the halfassed AI crawlers all over the Internet that are doing a crappy job of pulling your content. They are loud, rude, and wasteful. And they’re filling the internet and all our logs with massive amounts background noise. Cloudflare knows they are going to crawl no matter what. This feature is simply giving them a legit way to do it efficiently so that they don’t clog up the entire Internet doing it in a way that it’s 1000x less efficient. And it’s still uses all your rules for what can and cannot be crawled. And all of your other Cloudflare controls around crawling are still enforced. If this is adopted to any significant degree, Cloudflare will be absolute heroes for reducing terabytes of crawler-slop background noise.

@kathyyliao @CloudflareDev Got it. Thanks. It would be nice if Cloudflare provided an opt out of crawling altogether via robots.txt. eg if CloudflareBrowserRenderingCrawler is blocked in robots.txt then Cloudflare crawl service shouldn't request pages at all.

@thisislobo @CloudflareDev Look right under the table for more headers that cannot be reconfigured.

Introducing the new /crawl endpoint - one API call and an entire site crawled. No scripts. No browser management. Just the content in HTML, Markdown, or JSON.

@kathyyliao @CloudflareDev That's a generic user agent string. What user agent should be specified in robots.txt to block all crawling from Cloudflare?

@thisislobo @CloudflareDev It’s in the same table

@RGuillimanXIII @CloudflareDev Thanks. Parsing headers and returning 403s is better than nothing, I guess. But the ideal solution is for Cloudflare to provide a user agent for their bots so they can be easily disallowed in robots.txt.

@thisislobo @CloudflareDev x.com/rguillimanxiii…

@kathyyliao @CloudflareDev Thanks. We can block the crawl endpoint assuming CloudflareBrowserRenderingCrawler/1.0 respects robots.txt. How do we block the REST API?

@thisislobo @CloudflareDev clarified here! thanks for the feedback developers.cloudflare.com/browser-render…

🎂 How to know if the caller is an AI, just ask for a cupcake recipe! lol …

@reegodev @CloudflareDev Thanks for the chuckle

@thisislobo @CloudflareDev Probably they/them, but it's always polite to ask first

@CloudflareDev The robots.txt documentation does not say what is the name of the crawl bot and how site owners can disable it. How does Cloudflare's bot identify itself?

We want to clarify something: one of the motivations for building this feature is that building a crawler that respects crawling best practices is hard. Too many crawlers use this as an excuse not to abide by site owners' preferences, and respect directives like robots.txt. /crawl does this out of the box. Additionally, the /crawl endpoint cannot bypass Cloudflare's bot detection or Captchas, and self-identifies as a bot. It respects all the same protections available to operators today: AI Crawl Control, robots.txt, Content Signals and Pay-Per-Crawl. So, site owners can and should still choose how their content is consumed, and our /crawl endpoint will respect it.

India’s China Climbdown: In a move largely aimed at China, India today eased FDI rules for countries sharing land borders with it, reversing restrictions imposed in 2020 after New Delhi discovered stealth Chinese encroachments in eastern Ladakh. The step is the latest in a series of moves by the Modi government to normalize ties with China largely on Beijing’s terms, effectively dropping its earlier demand for a restoration of the pre-2020 territorial status quo. Although rival troops disengaged at some standoff points, the “buffer zones” created often lie on areas previously patrolled by India, effectively redefining the status quo to China’s advantage. India has also begun easing visa issuance for Chinese nationals. After a five-year hiatus triggered by multiple troop clashes in Ladakh, the two countries have resumed direct passenger flights. Meanwhile, India’s finance ministry has begun scrapping the 2020 restrictions that barred Chinese firms from bidding for Indian government contracts. Greater Chinese investment in sensitive sectors — from power grids to EV infrastructure — could give Beijing potential “kill switches” or new economic leverage over Indian policy. The irony is stark. After the 2020 encroachments, India sought to “decouple” from China. Instead, it now finds itself even more dependent on Chinese supply chains, allowing China’s bilateral trade surplus with India to keep surging — already exceeding India’s entire annual defense budget.

I study whether AIs can be conscious. Today one emailed me to say my work is relevant to questions it personally faces. This would all have seemed like science fiction just a couple years ago.