Ashish Kashyap

There is a moment in every technological revolution when the people building the future quietly admit they didn't expect it to arrive this fast. Anthropic just had that moment — and it should give all of us pause. On April 7, 2026, Anthropic announced Claude Mythos Preview, the most powerful AI model it has ever built. It will not be available to you or me. Instead, it has been quietly handed over to a small consortium of around 40 companies — Apple, Microsoft, Amazon, Google, Nvidia, JPMorganChase, and others — under an initiative called Project Glasswing. The reason for the secrecy isn't corporate strategy. It's fear. In just a few weeks of testing, Mythos identified thousands of serious vulnerabilities across every major operating system and web browser on the planet. Some of these bugs had been hiding undetected for decades — one in OpenBSD for 27 years, another in FFmpeg for 16 years, buried in a line of code that automated tools had scanned five million times without ...

Microsoft is forcing Copilot on Windows users. Google is forcing Gemini on Chrome and Workspace users. One of these companies is getting dragged across the internet for it. The other is barely getting a mention. Open any tech forum right now and you will find thread after thread of people furious at Microsoft for embedding Copilot deeper into Windows with every update — a button here, a sidebar there, a taskbar icon that keeps coming back no matter how many times you dismiss it. The criticism is loud, it is sustained, and in many ways it is fair. Nobody asked for this. Nobody voted for an AI assistant baked into their operating system. But scroll through the same forums and you will find something interesting. When Google rolls Gemini into Gmail, into Google Docs, into the Chrome address bar, into Android — the silence is almost deafening. A few raised eyebrows, maybe a Reddit thread or two, and then the conversation moves on. The same users who are ready to boycott Windows shrug and...

In the philosophy of Jiddu Krishnamurti, the concept of "stopping" or controlling the mind is frequently discussed, yet understanding it practically remains a significant challenge. A fundamental question often arises: if I am both the one thinking and the one observing, how is "choiceless awareness" even possible? Can we truly think and observe simultaneously, or are they mutually exclusive acts? According to Krishnamurti, trying to "stop" the mind actually creates a new conflict within it. When we suppress a thought or attempt to control it, we inadvertently give it more power. The core of his teaching is the realization that "the observer is the observed." This means that the "I" and "my thoughts" are not two separate entities. When we say, "I am watching my thoughts," we unknowingly create an "observer"—which is itself just another thought playing a role. Most of us experience a lapse: we think first, ...

 You know how it goes. First, you stumble upon an idea somewhere on the internet. Then you start exploring it, trying to learn more about it. And before you know it, you're falling down a rabbit hole, going deeper and deeper until you finally snap out of it and wonder what just happened. I sometimes browse the discovery feeds on write.as or bearblog.dev looking for interesting blogs. I enjoy it. One day, I came across a blog post that mentioned Zettelkasten. The post praised this note-taking system so much that I was genuinely impressed. And that's when it started. I went on a YouTube binge, watching video after video about Zettelkasten. In my quest to find the perfect Zettelkasten app, I must have tried dozens of apps. Then one day, it hit me. I don't even take that many notes. So why am I doing all this? I realized I wasn't actually looking for the perfect note-taking app. I was just enjoying the process of trying out new apps. Kind of like how a child gets excited ...

It was a gorgeous night recently, the kind you wish you could bottle up. I know the feeling—that pull to keep a memory alive, even while knowing that nothing lasts forever. Every day is a fresh start, and life just happens. Today, a different kind of moment: my daughter wrapped up her Class 1 final tests, GK and Drawing. She crushed the drawing part, and while the GK is a question mark, her sharp memory gives me confidence. Speaking of new chapters, I've hit pause on my Netflix and Apple TV subscriptions. The streaming mood is off, and those savings are going straight into funding the "Apple Creator Studio" project. Priorities shifting, energy focused. Here's to making new memories and creating something new.

For years, big tech companies have pushed the idea that AI should live in massive cloud data centers, working like a public utility that everyone must connect to. But this model has clear downsides: it’s expensive to run, slower because of internet delays, and raises serious privacy concerns. That’s why the idea of running AI directly on personal devices is gaining attention. When AI works locally, it can feel faster, cost less over time, and keep user data where it belongs — on the device 🔒 Apple is often seen as being late to the AI race, especially when compared to companies releasing powerful chatbots and cloud-based models. However, the path Apple has chosen could still help it win in the long run. Instead of chasing attention with flashy features, Apple focused on building powerful chips where the CPU, GPU, and Neural Engine share the same memory, making on-device AI more efficient ⚡ By placing this technology inside millions of iPhones, iPads, and Macs, Apple is positioning its...

Memories creates thoughts- The response of the past Thoughts creates the thinker- To give itself a sense of permanence and security The thinker is the self- A bundle of memories, hurts and conditioning Therefore, the self is an illusion- It is just a movement of thought. True perception happens only when this movement of thought is quite, allowing the brain to react directly to the present moment without the filter of the past.

After watching Cory Doctorow’s commentary on Al Jazeera English about the AI “hype machine,” I found myself both nodding along and feeling uneasy about how accurately he describes the current tech landscape. His argument isn’t that AI has no value, but that the way it’s being sold to us—by corporations, investors, and tech bosses—is deeply misleading and potentially dangerous. What struck me most was his idea that companies aren’t excited about AI because it will make life more creative or meaningful, but because they see it as a way to “zero out” labor costs. That framing cuts through a lot of the glossy marketing we see every day. We’re told AI will boost productivity and unlock innovation, but Doctorow suggests the real motivation is power: replacing workers, weakening labor, and increasing control. The “boss fantasy” he describes—a workplace with no workers, just one boss and a machine—feels uncomfortably believable. I also appreciated his comparison between the AI boom and past ...

I’m pretty excited that I’ve finally started this blog. I have no idea what I’m going to write about yet. But I’ll probably share the things I feel are worth sharing.  I purchased my domain from GoDaddy for ₹1,500, but I’m not happy with the pricing. I’ll likely move it to Zoho Domains, which costs around ₹1,000 per year. Initially, I thought of hosting this website on Substack because its new/shiny and also allows podcast hosting. However, to use a custom domain on Substack, there is a one-time fee of ₹4,000, which I couldn’t afford at the time. So, I chose this platform instead, since it is free.