periwinkle.eth

Expectation: the age of the IDE is over Reality: we’re going to need a bigger IDE (imo). It just looks very different because humans now move upwards and program at a higher level - the basic unit of interest is not one file but one agent. It’s still programming.

New in Claude Code: Remote Control. Kick off a task in your terminal and pick it up from your phone while you take a walk or join a meeting. Claude keeps running on your machine, and you can control the session from the Claude app or claude.ai/code

Despite thousands of listings that appear to have illegally raised rents since the January fires, L.A. County prosecutors have yet to file any charges. Now, county leaders want a new approach to enforcement: fines of up to $1,000 per day. Read our @LAist story for details 🔗👇

I’m hiring a product designer for Sudowrite. This is a dream job for a designer who loves to write. The beginning of the listing is… a mini therapy sesh for me. You’d think it’d be chill after that. But no, it goes on to talk about our company cult… sudowrite.com/jobs/designer

This is peak AI 😂

Here's my thread about why you should: - quit whatever it is you're doing - build the startup you've wanted to build - work with me and the a16z team 1:1 - And yes, I'll invest $1M via a16z speedrun (and the earlier your startup, the better. apply here: sr.a16z.com) I want to start with some things I hear: "Nah, it's not the right time for me to launch a new startup. But why yes, I do scribble down new ideas all the time. And my best friend from work would be the perfect cofounder. Oh I even have a side project going..." There's always something around the corner: - a promotion that might happen - a job you want to get before you're "ready" to start a company - the perfect first hire that might free up in a month - a work project you want to finish up first - a sign that your idea is the perfect idea ... and a million other things But let me make the counterpoint, which is that the AI has created a unique window for launching new products. And it's ephemeral and (unfortunately) rapidly closing. - the startup window opens and shuts for a few years, once every decade. Wait it out now, and you'll have to wait 10+ years for the next one. After all, the AI wave is huge and comparable to what we saw in mobile, internet, desktop, etc. And if you look at the dates, you'll see -- 2007 for mobile, 1994 for web, 1984 for desktop. These waves happen sparingly. You could build Instagram in the first few years of mobile, but you don't want to be the photo-sharing app 8 years in. You could build Uber in the early days, but don't try to compete with them now. These windows are key. - the excitement for trying new products is at an all time high. On social media, we've now all seen first hand when new products are launched and hit great traction right off the bat. People have a hunger to try new products in a way that hasn't existed since the App Store launched. - one of the biggest problems for pre-AI products is that marketing channels are incredibly stagnant. Mobile ad channels are expensive, SEO is slow (and Google keeps taking over the search page), and you can't use aggressive virality anymore. Growth is hard. But if you have a new AI-driven take on a product category -- take advantage of the novelty effect! Before the marketing channels get swamped. - adding AI into an idea no longer requires a PhD in Computer Science :) This means that you can build totally wow breakthrough features into your product and impress your users simply by integrating APIs. You don't need to train any models, and your users may not even care what's under the hood. (just look at cursor, granola, etc -- they just care about great UX). The infrastructure is largely here, but the products are still early. There are still tons of great opportunities. - there's an incredible talent arbitrage right now. There aren't that many AI-native jobs, but everyone wants to learn more, so you can recruit top people if your product starts life day 1 as an AI product. - you might ask, can't I just work on this next year? Well -- maybe. These big waves are intensively competitive. The good ideas are getting snatched up right now, as people ask the question for how to apply AI to XYZ sector. And once the area is taken by a good team with real funding, you could maaaaybe fast-follow, but you're playing on hard mode at that point - there's never been a better time! After all, I found this fun post a while ago: Building a startup in 1995: - Buy & run a server - Code everything from scratch - Handle payments manually - Market with print ads Building a startup in 2025: - Click to deploy - AI writes code for you - 1 line of JavaScript to charge any credit card - 1 tweet goes viral = thousands of visitors (h/t @marclou) - another good reason to get started: You're reading the right essay :) And now you know about a16z speedrun, which makes it possible for us to invest in very early startups, particularly in our favorite area of tech and entertainment and AI. This is still an experimental program, and we're committed to invest up to $1M in teams that are even pre-product and/or pre-traction, if you've had strong relevant experiences. Things are early enough that I am super hands on, and look forward to working with many of you! And the link again: sr.a16z.com and more program details: speedrun.a16z.com Ultimately, a big wave like this doesn't come often. It's easy to make excuses, and for those of you who know you want to be a founder one day, I encourage you to consider jumping in during 2025. - For those of you who want to wrap something at work, or are waiting for a promotion, or something else, I want to tell you a secret: No one will care, after you've spent 18-24 months+ in a single job, what happens next at that job. The tech industry changes quickly and you get rapidly diminishing returns from what you can learn/prove at a single place. You can cut it short, and jump into a new startup. - For those who are waiting for the perfect idea: Your first idea will probably suck. Well maybe that's too harsh, so the more polite way to say it would be that it's based on incomplete information. You don't have a product in front of real users and customers. You haven't explored all the competitive products. You haven't spent 1000s of hours talking to your users yet. You need to get started doing that, to get through the 100 variations of your sucky ideas to get to the real one. - For those who are waiting for a sign: Here's your sign :) - For those who don't have an idea: That's OK. Apply with some areas/sectors you are interested in, and we'll help you talk through the kinds of problems that are exciting - For those who want to work on another job before you know enough to start a company: You know this isn't logical. Plenty of founders get going before they have any significant work experience. Another set of founders often start companies in sectors where they have no prior expertise. You don't need more prep!

I heard this internal memo of mine is being leaked right now, so here it is:

I 100% vibe coded a macOS app (Cocoon) with Gemini 2.5 pro, Claude Sonnet 3.5 and 3.7, including landing page, all READMEs and even this announcement itself! Check it out & download: 🔗 Landing Page: amol-patil.github.io/cocoon/ 🐙 GitHub: github.com/amol-patil/coc… 1/

Moment of the 7.7 Magnitude powerful earthquake in Bangkok, Thailand 🇹🇭 (28.03.2025)

Market Thoughts No sugarcoating it, the market is brutalizing everything but Bitcoin atm. BTC down to 95k with BTC Dominance at 61% is a sight to behold - basically everything bleeding against it, and bleeding massively against their USD pairs. There's an insane amount of fud going around now -- I think sometimes people forget what fud stands for: fear, uncertainty, and doubt. This is obviously because of Trump and the tariffs, but also just general market weariness and hesitation over everything Trump related. It's hard to say how warranted it is, but whenever I think about it, I realize that nothing fundamentally has changed. He is doing everything he said he would. He has followed through on all promises so far. One of the things he said is that he would be the "crypto president", and we have a tonne of positive changes already made: ross freed, SBR proposed, gensler gone, crypto council/czar, etc etc. The medium and long term outlook for crypto is better than it's ever been. The short term is anyone's guess, as it's always been. The best thing to do in times like this is to take a step back, breathe, and evaluate. Think about what is within your power to change, and if there is anything you can learn. Were you too heavy in alts? memecoins? ai agents? nfts? Did you have 80% of your portfolio in the riskiest assets and little to no majors? no stables? You can't change the past, but you can learn from it and make better decisions going forward. I've made a lot of mistakes myself. I was probably too heavy in all of the above categories. Even though I took more profits over the last year than ever before in my crypto journey (both in absolute terms and in % terms), I still think I probably ought to have taken more and been more consolidated into the majors, and less spread-too-thin. Can't change the past. But will use this as a reminder going forward to try and do better. In terms of specific holdings, I remain bullish on a lot of the things I have been bullish on these last few months: PENGU, "nothing token", AIXBT, VADER, VIRTUAL, HYPE, PRIME. And ofc the majors, max bullish on ETH still, and also bullish on SOL and ofc BTC. If the bull market continues (I think there's a v high chance it does), these should all bounce back strongly imo. Yes there are an infinite number of tokens out there, but there aren't an infinite number of good tokens. That's the difference. There are a lot of them, sure, but it's relatively easy to separate the junk from the legitimate and the tokens that have been around for a while, have teams and communities backing them, stand a strong chance of coming back. This is no guarantee of course, and the reason for spreading bets into a few is that chances are at least one of the above mentioned tokens won't recover. But the r/r is that the gains from the one(s) that do, will make up for the losses in the other. I know it's scary out there and it's not fun to log in every day and see your portfolio down another 20-30% every fucking day. This is crypto though. This is the insane world we choose to partake in. Tremendous gains would not be possible without the risk of tremendous losses. It's why I always say the best strategy for most is simply to DCA into the majors and hold and wait a bunch of years. Most people here choose not to do that though (myself included), and that's fine, but we must acknowledge that we're playing with fire and accept that we're gonna get burned from time to time. Don't let it get you too down -- you're still in a better position than most, and you're in roughly the right place, at roughly the right time, surrounded by opportunities. Infinite Regret and all that. And if all else fails, being grateful for health, friends, family, pets, food, roof, etc etc can help mollify the pain. 💜

@SavageSausageXY @RT_com source?

@RT_com 5% containment.

No stopping it: Hughes Fire approaching 10,000 acres — zero percent containment with evacuation orders in place

Introducing Willow, our new state-of-the-art quantum computing chip with a breakthrough that can reduce errors exponentially as we scale up using more qubits, cracking a 30-year challenge in the field. In benchmark tests, Willow solved a standard computation in <5 mins that would take a leading supercomputer over 10^25 years, far beyond the age of the universe(!).

$META's iPhone moment

It’s official…programming as a career is over. GPT o1-preview creating a fullstack react native app…I didn’t make a company fast enough, and now I’m obsolete…fullstack for 16 years

@EthereanVibin ETH bullposting is good

Hard to stay mad at Apple. Without their turn to monopoly bully, I probably wouldn't have discovered the joys of Neovim, Zellij, or Linux on the desktop. I would have slept on the amazing Framework 13. Omakub would not be. Amor fati and thanks for the push, Cupertino 🙏

this is a cool one! we will improve the feature quickly.

Lily Zhang’s parents get a Gold medal in Asian Parenting

Your 9-to-5 job is dying. By 2034, it'll be extinct. That's Reid Hoffman's latest prediction – the founder of LinkedIn who predicted the rise of social media in 1997. Here's what he said next:

This was one of the sketchiest parts of Las Vegas 25 years ago