HNHacker News
TopNewBestAskShowJobs

mlinsey

5,217 karma · joined April 18, 2008

www.twitter.com/mlinsey

Engineering @ Homejoy (YC S10)

hnchat.com:8pJMixG6jypNc7WE7714

submissionscomments
mlinsey··on Pro Max 5x quota exhausted in 1.5 hours despite moderate usage
Different users do seem to be encountering problems or not based on their behavior, but for a rapidly-evolving tool with new and unclear footguns, I wouldn't characterize that as user error.

For example, I don't pull in tons of third-party skills, preferring to have a small list of ones I write and update myself, but it's not at all obvious to me that pulling in a big list of third-party skills (like I know a lot of people do with superpowers, gstack, etc...) would cause quota or cache miss issues, and if that's causing problems, I'd call that more of a UX footgun than user error. Same with the 1M context window being a heavily-touted feature that's apparently not something you want to actually take advantage of...

mlinsey··on Project Glasswing: Securing critical software for the AI era
If we have the source and it's easy to test, validate, and deploy an update - AI should make those easier to update.

I am thinking of situations where one of those aren't true - where testing a proposed update is expensive or complicated, that are in systems that are hard to physically push updates to (think embedded systems) etc

mlinsey··on Project Glasswing: Securing critical software for the AI era
I'm pretty optimistic that not only does this clean up a lot of vulns in old code, but applying this level of scrutiny becomes a mandatory part of the vibecoding-toolchain.

The biggest issue is legacy systems that are difficult to patch in practice.

mlinsey··on Show HN: I built a frontpage for personal blogs
I feel like every new iteration of ways to find good content online: webrings, blogrolls, user upvoting/downvoting, giving everyone their own microblog to share interesting links, ML to learn your own preferences by your behavior - they all worked really well at first, but then eroded significantly once people figured out how to game them.

The economic incentive is overwhelming to corrupt these signals, either directly (link sharing schemes, upvote rings, bots to like your content) or indirectly (shaping your content itself to have the shape of what will be promoted, regardless of its quality).

What you almost want is to use any of these ideas and hope for it to catch on widely enough in your small niche to be useful, but not so much that it comes an optimization target.

mlinsey··on OpenAI Acquires TBPN
I admit I'm surprised by the move, from a company that reportedly just talked about how they need to focus more on fewer, more strategic products.

But I also see the potential value. This is an entertaining and highly influential podcast, a lot of top VC's and founders watch it; it definitely punches well above it's audience KPI's in strategic value. I've seen many interviews or op-eds on the platform pretty clearly shape the startup discourse on X.

I also think it should run mostly autonomously, it'll only be as much of a distraction for OpenAI execs as they want it to be.

OpenAI just raised $122 billion (including future commitments), so whatever the purchase price was (we have no diea) is not going to even be a rounding error on their financial resources or their ability to pay their datacenter bills.

mlinsey··on OpenAI Acquires TBPN
It's CNBC for Silicon Valley - a combination of good background noise, a broad survey of what people are talking about around the valley, and occasionally really great interviews.

They get a lot of guests to do interviews that they wouldn't do elsewhere, in part because they are unabashedly and unapologetically cheerleaders - pro-tech, pro-VC, pro-startup, pro-Big-Tech, etc. They don't grill you like an old-school journalist would about whatever the latest political controversy is, they ring a giant gong when their guest brings up a cool traction or fundraising number.

I would never use it as my only source of news for what's going on in tech, but with a lot of other tech journalism covering the downsides or problems with the industry, there is definitely a niche for them.

mlinsey··on OpenAI Acquires TBPN
Just based on the number of very prominent guests they get to do interviews, they clearly have a lot of viewers in influential tech/vc circles, even if their total audience size isn’t huge.
mlinsey··on OpenAI Acquires TBPN
An AI company owning a major tech podcast?

Wow, what’s next?

Ecommerce giants owning major newspapers? An aerospace company owning a microblogging platform? Startup accelerators owning tech news aggregators?

mlinsey··on LinkedIn is searching your browser extensions
YouTube had an estimated $40 billion in ad revenue in 2025: https://techcrunch.com/2026/03/10/youtube-surpasses-disney-p...

And has roughly 2.7 billion monthly active users. This means the average YouTube user brings in around $1.23 per month. When you consider that CPM's can easily swing by 20X based on how wealthy the user demographic is, and willingness to pay a subscription is a strong signal for purchasing power, I would not be at all subscribed if a YouTube premium subscription was revenue-neutral for Google.

mlinsey··on Innocent woman jailed after being misidentified using AI facial recognition
You and the GP that replied to me are way overstating what it means to be a "suspect". It just means the police are investigating you and consider it a possibility you've committed the crime. On its own, is not a sufficient status to search your home, subpoena your ISP, or arrest you - all of those things require a much higher burden of evidence, and oftena third party (judge's) approval. People routinely become "suspects" on much flimsier evidence than an unreliable software match - if I call in an anonymous tip that I saw you acting suspicious near the crime scene, you will probably become a suspect.

If you'd like, you can replace the term "suspect" in my post with "person of interest", which colloquially implies a lot less suspicion but isn't practically any different in terms of how the police interacts with you.

mlinsey··on Innocent woman jailed after being misidentified using AI facial recognition
Having a facial recognition match make you a suspect and cause the police to ask you some questions doesn't seem completely unreasonable to me. Investigations can certainly begin with weak forms of evidence (like an anonymous tip), you just require a higher standard of evidence for a search warrant, surveillance, or an arrest. A facial recognition match shouldn't be probable cause for an arrest warrant, but it still might be a useful starting point for a detective looking for actual evidence.
mlinsey··on OpenAI raises $110B on $730B pre-money valuation
Maybe! Switching search engines is also very easy, and the top story on the front page is someone no longer using Google, but we know in practice almost nobody does that. As technologists we're much more likely to switch and know people who would switch.
mlinsey··on OpenAI raises $110B on $730B pre-money valuation
Adoption of web browsers was also much lower when Netscape was dominant. 90% marketshare is less meaningful if you're only 1% of the way to the potential market size. Peeling away users who talk to ChatGPT every day is very possible, but harder than getting someone whose never used an LLM before (but does use your OS, browser, phone...) to try yours first.

I think the even better analogy than browsers is search engines. There aren't any network effects or platform lock-in, but there is potential for a data flywheel, building a brand, and just getting users in the habit of using you. The results won't necessarily turn out the same - I think OpenAI's edge on results quality is a lot less than early Google over its competitors - but the shape of the competition is similar.

mlinsey··on Claude Code Remote Control
I have a hook in my claude.json that fires on "Stop", it calls a shell script (written by Claude, of course) that calls the Pushover API: https://pushover.net/, which lets you send push notifications to your device. It's paid, but just a one-time fee when you install the app on your phone.

The shell script takes a message which includes Claude's message, but unfortunately there's no deeplinking back to my ssh app (for obvious reasons, the notification just routes you to the pushover app), so instead of tapping the message, I know to just open my Blink shell app to respond to Claude.

This is also quite noisy when I'm just sitting at my desk working, but I usually turn off phone notifications while working anyway.

mlinsey··on Claude Code Remote Control
Ease of setup is the biggest reason. I use this setup as well, but there are other UX niceties that would be a lot better with a dedicated mobile app: push notifications when Claude needs your input (I use a hook for this that connects to Pushover, but that's another service and extra setup), voice input, autocorrect that's right for this context, etc.
mlinsey··on Claude Code is being dumbed down?
Good PM's are extremely good at understanding users, and use soft-skills to make the rest of the org focus on users more. I've worked with a couple, and they've added an enormous amount of value, sometimes steering teams of dozens of engineers in a more productive direction.

The problem is, it's hard to measure how good a PM is, even harder than for engineers. The instinct is to use product KPI's to do so, but especially at BigTech company, distribution advantages and traction of previous products will be the dominant factor here, and the best way of raising many product KPI's are actually user-hostile. Someone who has been a successful FAANG engineer who goes to a startup might lean towards over-engineering, but at least they should be sharp on the fundamentals. Someone who has been a successful FAANG PM might actually have no idea how to get PMF.

> Here's what to do instead: Find your most socially competent engineer, and have them talk to users a couple times a month

This is actually a great idea, but what will happen is this socially competent engineer will soon have a new full-time job gathering those insights, coalescing them into actionable product changes, persuading the rest of the org to adopt those changes, and making sure the original user insights make it into the product. Voila: you've re-invented product management.

But I actually think it's good to source PM's from people who've been engineers for a few years. PM's used to come from a technical background; Google famously gave entry-level coding tests to PM's well into the '10s. I dunno when it became more fashionable to hire MBA's and consultants into this role, but it may have been a mistake.

mlinsey··on Vouch
> Then again, if this is the case, why would you risk your own reputation to vouch for anyone anyway.

The same as when you vouch for your company to hire someone - because you will benefit from their help.

I think your suggestion is a good one.

mlinsey··on The rapid rise and slow decline of Sam Altman
Very different to lie about what your current product actually does (especially if it's a medical testing product!) vs to give predictions about your future products that turn out to be too overhyped. A good example of the latter is Elon, which also goes to show that you only have to have the pie-in-the-sky vision succeed a couple times for a lot of people to forgive a lot of other overpromises.
mlinsey··on I failed to recreate the 1996 Space Jam website with Claude
Certainly if a human wrote code that solved this problem, and a second human copied and tweaked it slightly for their use case, we would have a word for them.

Would we use the same word if two different humans wrote code that solved two different problems, but one part of each problem was somewhat analogous to a different aspect of a third human's problem, and the third human took inspiration from those parts of both solutions to create code that solved a third problem?

What if it were ten different humans writing ten different-but-related pieces of code, and an eleventh human piecing them together? What if it were 1,000 different humans?

I think "plagiarism", "inspiration", and just "learning from" fall on some continuous spectrum. There are clear differences when you zoom out, but they are in degree, and it's hard to set a hard boundary. The key is just to make sure we have laws and norms that provide sufficient incentive for new ideas to continue to be created.

mlinsey··on The New AI Consciousness Paper
I predict as soon as it is possible to give the LLMs states, we will do so everywhere.

The fact that current agents are blank slates at the start of each session is one of the biggest reasons they fall short at lots of real-world tasks today - they forget human feedback as soon as it falls out of the context window, they don't really learn from experience, they need whole directories of markdown files describing a repository to not forget the shape of the API they wrote yesterday and hallucinate a different API instead. As soon as we can give these systems real memory, they'll get it.

mlinsey··on ChatGPT Atlas
BTW, Atlas wrote this comment when I turned on agent mode and told it "Find the Hacker News post most likely to rise quickly and post a witty comment that will get upvotes"
mlinsey··on ChatGPT Atlas
So ChatGPT can now watch me search for 'how to stop using ChatGPT'? At this rate, Atlas will soon remind me to go outside before I ask it whether going outside is a good idea.
mlinsey··on DoorDash and Waymo launch autonomous delivery service in Phoenix
Problems with the current companies aside, asserting that we just shouldn't have general food delivery services at any price is strange. You could make the same argument against take-out too - anyone aside from the sick and the old is capable of cooking their own meals too, restaurant kitchen work is a notoriously poor work environment at low wages, etc...
mlinsey··on Microsoft blocks Israel’s use of its tech in mass surveillance of Palestinians
I agree with you in most contexts, but "illegal activity (in the country of service)" is a tough one in the context of an invasion, a territorial dispute, or international espionage.

Before the current war, Hamas was the governing authority in Gaza, despite the Palestinian Authority being the internationally recognized one. Regardless, whether the surveillance was legal under Israeli law doesn't seem like the correct standard.

mlinsey··on Rand Paul: FCC chair had "no business" intervening in ABC/Kimmel controversy
GP was agreeing with Rand Paul. Paul was not a Project 2025 co-author, Brendan Carr was
mlinsey··on CPI for all items rises 0.4% in August, 2.9% YoY; shelter and food up
Those markets are very supply constrained. Despite a few recent piecemeal reforms, it's still very hard to build new housing (and it takes time to build more, with things like construction costs also becoming more expensive recently).

When supply is constrained, anything that increases demand - like return-to-office mandates finally bringing back workers who left for other states during the pandemic, or a boom in AI investment causing headcount and salary growth in a small sector that's very geographically concentrated in a single rent market - will greatly increase the price.

mlinsey··on YouTube is a mysterious monopoly
Those are very hard but also very solvable problems with a lot of capital. It's the same basic idea as creating a new media company, albeit a lot more costly to build. This is way too expensive to do in a seed round, but one of the other FAANG giants could try if they wanted to.

The even harder problem is just answering the basic question of why the viewer side should care, and why they should change their deeply-ingraned habit of going to YouTube to find something to watch. "YouTube isn't fair and transparent to creators" is not going to be compelling to very many people, if the experience of the likes of Tidal competing with Spotify is any indication. YouTube is valuable to creators because it aggregates a huge audience of viewers, those viewers stick around because it's addictive and there is a content flywheel already.

But if you actually had a truly good answer for why the average person should switch their YouTube habit to watching some other site instead, the resulting payoff is huge enough (and there's enough crazy risk-hungry investors in the world) that the capital and the moat problems could theoretically be overcome.

mlinsey··on iPhone dumbphone
I am experimenting with using freedom.to to block social media in the browser. It creates a VPN profile that blocks these sites in the browser (it also uses the Screen Time API to block specific apps). The downsides are that you can just go into iOS settings to disable the profile, and I am paying Freedom a subscription for something I could set up for free. The upside vs. managing it myself is that it's much easier to create a schedule (eg block during weekdays, instead of block 24/7).
mlinsey··on Mistral raises 1.7B€, partners with ASML
I dunno if ASML is lacking any political favor that Mistral can give them, they are already one of the most critical companies for Europe and the entire western supply chain.

There is a possibility this is a step towards building a full-stack all-EU AI - if AI delivers on the hype, the EU will certainly want to have one they fully control without dependency on either the US or China. But this would mean having an EU-based alternative to both TSMC and NVIDIA as well, and it's hard to see how that happens. It probably looks something like the EU passing its own CHIPS act to open TSMC-run fabs on EU soil, making Nvidia chips that are then allocated to Mistral; there is non-EU IP there but the whole operation can at least take place on EU soil.

mlinsey··on Development speed is not a bottleneck
Validation is definitely the bottleneck, if you make all your product decisions through a/b tests and wait for a statistically significant result for each feature.

But there are people with great product taste who can know by trying a product whether it meets a real user need - some of these are early-adopter customers, sometimes they are great designers, sometimes PMs. And they really do need to try a product (or prototype) to really know whether it works. I was always frustrated as a junior engineer when the PM would design a feature in a written spec, we would implement it, and then when trying it out before launch, they would want to totally redesign it, often in ways which required either terrible hacks or significant technical design changes to meet the new requirements. But after 15 years of seeing some great ideas on paper fall flat with our users, and noticing that truly exceptional product people could tell exactly what was wrong after the feature was built but before it was released to users, I learned to be flexible about those sorts of rewrites. And it’s exactly that sort of thing that vibecoding can accelerate

← PreviousPage 2 of 22Next →