Will customer willingness to deal with fast-moving products become the next bottleneck in software? Spoke with someone redesigning the SDLC at an infra company that many of you use They’re big Amp fans and she wanted to learn more about how we’ve changed they way we work to accommodate agents She mentioned Goldratt’s Theory of Constraints, pointing out that last year the bottleneck was code gen, this year it’s code review (although she’s doing away with that soon), and the next one she’s seeing: ** Her customers are increasingly frustrated by how quickly their product is evolving and new features are being released ** Someone in this situation might ask themself, shouldn't we slow down and listen to our customers? Isn't the customer always right? The reality is -- software companies are being forced to act this way because of competitive pressures If you move slowly and build only what your customers ask for, you'll be leapfrogged by an agent-native competitor that is focused on building the best possible product And so you find yourself in a dilemma. Competitive pressures demand that you keep shipping, but your current customers -- especially the really big ones that pay all the bills -- are asking you to slow down Which leaves you with a choice: - slow down, milk those big contracts for all they're worth, and hope an agent-native competitor doesn't leap-frog you - speed up, do damage control with those big customers, and try to rebuild your business around a new type of customer that expects you to turn around features and bug fixes in days/hours rather than weeks/quarters Choosing the first is implicitly a bet against AI progress and diffusion into the real economy Because you know what causes your customer to start demanding agent-native shipping velocity? When they themselves are feeling nigh-existential competitive pressures

If you're looking for a Conductor replacement, it's @AmpCode It handles my: - Customer support - SEO via - GitHub issues - iOS apps (building, submitting, ASO) - Product features + planning And I use Puck as a personal assistant. Just Orb it.

Leaving Conductor? You can bring your chats to Amp. 1. Update Amp 2. Run this in your terminal: `amp threads import conductor --all` Using the Amp Mac app? Open Settings, go to Imports, and click Import. Chats in Conductor Cloud: move them to your Mac first (right-click → Transfer to local) before Nov 20. Once you're imported, use the following code to get 3 months Megawatt plan on us:

There used to be a convenience store one block from my house that I’d walk to a few times per week for things like batteries or paper towels or a random ingredient we were missing for a recipe That store closed and it’s been replaced by an Amazon warehouse sort of thing. The windows are covered but there are signs on the door and 10-12 deliver drivers in Amazon vests milling in and out all day long I like to imagine that it’s stocked with the same convenience items as before but now you’re intermediated by a delivery driver and a 24-48 fulfillment time Weird times

okay public speaking is hard, apologies for the folks who showed up to watch me overheat and bomb, so here’s all the stuff i meant to explain right now you are being asked for a near 100% prepayment because neither you nor the cloud is considered credit worthy, and so nobody can get debt. As ai scales, we are asking the credit markets to take on a massive single sector deployment of capital. More money deployed, more scared everyone gets. this problem will flow down into the token layer. Every inference provider or model lab struggles to be credit worthy enough to get compute, and their lever to improve their situation is to lock in big prepayments from their customers or long term contracts from credit worthy customers. So this problem doesn’t just effect compute buyers, it effects token buyers. We’re entering a world where fortune 500s can very quickly deploy huge token spend ON CREDIT where the startup will need to raise capital. Startups got to exist because human capital could leave a firm and get ahead of the big firm by exploiting a combination of incentive misalignment, speed, and taste. You probably have the hidden assumption that the fortune 500 is slow and incapable of competing with your ultra fast startup. This belief does not necessarily hold in the case where the agent does the thing. the “market” is working on solving it. We know, because we are the market, and we’re working on it. sfcompute is working on ways of firming up the collateral of the loans, which makes it safer to lend to a startup. this is basically the whole reason we exist, the problem everyone is experiencing now is the problem junelark, our little ai lab years ago, hit however, we won’t succeed in time. so, in the short term, startups and venture firms should learn to protect themselves. we are mostly watching startups and investors make unforced error after unforced error, because this is not a simple subject. first, venture firms need to understand that they are in the compute business now and deploying capital to pay for a 100% prepayment for 10 startups where 6 of them die and 4 of them make 50% margins on that capital is basically locking in non-venture returns. many funds, via low understanding of the space, are wrecking their funds. there’s an economic solution for vc’s and it is owning the compute and trading equity for gpu/hrs. If the company dies, you keep the residual value. Each of you are looking to capture the next anthropic or openai, but the only way these companies scaled was by trading compute for equity. meanwhile, startups have three options in a credit crunch. The first is to raise a wild amount of capital, enough to pay for your deployments with far larger prepayment or to survive until things chill out. theres a reason we’re seeing a number of multi-billion dollar rounds simultaneously. The second is to design your business model to support variable priced tokens and compute. As the financial ecosystem builds around compute, you will be able to trade “lower up front payments” for “variably priced compute or tokens”. Your business will get tailwinds from the infra under you if you setup your business model such that YOUR customers have variable prices and you’re just adding a % take rate on top of the underlying floating price. Every year, you will become more and more capital efficient. The economic winners in ai & inference are firms that make lower and lower margin every year, but become more capital efficient & scale faster every year, making their gross profit increase. The third option is asynchronous or interrupted products. If you have a variable cogs, but a fixed rate card for your customers, it can’t have 100% availability or you’ll go bankrupt. But it is totally possible to design your product such that customers like this.

My wife during the workday, usually: won't talk to me for hours at a time, totally zoned in on her spreadsheets My wife during the workday, as I approach the end of a minutes-long prompt dictation: bursts into my office and starts talking about how cute one of our cats is

The so-called death of software had a false-start earlier this year, but there are now two additional forces pushing in this direction: - security teams locking down vendor-hosted saas sprawl - cloud agents Over the past few weeks, we've seen a huge increase in security incidents. Everyone in-the-know recognizes that we're still in the early innings here, things will get very bad over the next 12-24 months This increases demand on security team resources, significantly. The bar for getting through security review as part of the vendor onboarding process will be very, very high -- perhaps 5-10x harder than before Simultaneously, we're seeing the rise of cloud agents, which companies are typically self-hosting in their VPC. A clone of a much-needed vendor tool is now one cloud agent thread away Hand Fable an ephemeral VM, the docs of the vendor you're considering, and let it cook overnight with Opus doing implementation The next step, obviously, is easy self-hosted deployments of these clones. It'll take a few months for the right patterns to emerge, but someone will nail it, then everyone will copy, and by February you'll have Jason Calacanis raving about "self-hosted cloneware" on the All-In Podcast

Funniest thing I heard this week was a member of the team being surprised by how cool the people we met at last night’s NYC meetup were compared to typical tech meetups. “There were people drinking martinis!”

I’ve much preferred OpenAI models since ~Feb But since the launch of Astra, I’ve experienced both Astra and Sol behaving in ways we haven’t seen from frontier models since Sonnet 4 circa one year ago