z0.ai | Full Time | Founding Engineer | On Site 5 days a week | NYC | $140-$180k USD + equity
We're helping non technical users adopt AI by making usage visible. Power users become role models and people can upskill to gain value rather than lose their jobs to AI.
It's a complex problem because we need to balance trivial UX, with a slew of security and compliance requirements.
This is the first non founder developer role on the z0.ai codebase. We need someone who can own the full stack end to end: product surfaces people enjoy using, services that stay correct, and the taste to make both feel like one product.
We're funded and are live with our first customer already. We expect to onboard our next customer later this week and are actively ramping up sales efforts.
I have a hard time with the idea that Wikipedia is unbiased when the main source in most cases is news reporting. Wikipedia is a societal form of Gell-Mann Amnesia.
This is not entirely fair. The overwhelming majority of Wikipedia is not meaningless politics, but stated facts backed by decent sources.
The phenomenon you are referring to usually happens in areas where there is ideological or political friction. Sure, some articles can be biased, because staying perfectly factual in the middle of an active political debate or social change is difficult for most people. But in that case, there is still the option to edit the page or start a discussion.
If something is created by a community and editable by anyone, then yes, you can safely assume that certain topics will not be perfectly unbiased. But the fact that you can see the sources, edit history, and discussions that led to a given decision is already a major advantage.
Personally, I do not know a better alternative. I have a friend who told me Wikipedia is biased, so he refuses to use it. When I asked him what he uses instead, he said, completely seriously, “X is my main source of information.”
> But in that case, there is still the option to edit the page or start a discussion.
Honestly, I think on any politicised topic, that’s a waste of time - there’s a large contingent of Wikipedia editors with a shared deeply ingrained perspective that will reliably back each other up. There are better uses of one’s time than fighting such a losing battle.
> Personally, I do not know a better alternative. I have a friend who told me Wikipedia is biased, so he refuses to use it. When I asked him what he uses instead, he said, completely seriously, “X is my main source of information.”
I tend to use AI to surface sources and concepts, and then go read the sources for myself to verify the AI’s claims. AI has a strong tendency to e.g. misrepresent what journal articles say, but (if they are open access or otherwise available-and they generally are if an AI is citing them) you can then read them yourself and make up your own mind.
AI has genuinely taught me things I didn’t know before about topics of interest to me-e.g. Islamic history-but I’m careful to verify its claims with reliable sources rather than just trusting them-which of course one should do with Wikipedia too
>I have a friend who told me Wikipedia is biased, so he refuses to use it. When I asked him what he uses instead, he said, completely seriously, “X is my main source of information.”
I guess that was a few years ago? Because now he also has Grokipedia ("from the guy that brought you X")...
It was few years ago indeed. I honestly do not give Grokipedia much credibility because it was created solely after Musk got political and someone edited wiki saying that he aligned with "right" or "far-right" politics. He saw that and created it as an "facts based" alternative purel out of spite.
> Musk is positioning Grokipedia as an alternative to Wikipedia, which he called "Wokepedia" in an X post last December.
> Grokipedia also says Wikipedia is the subject of "persistent criticisms regarding factual reliability, susceptibility to vandalism and hoaxes, and systemic ideological biases — particularly a left-leaning slant in coverage of political figures and topics.
...which is consistent with what the right side of the US political spectrum keeps saying about media outlets that dare to disagree with them.
> The overwhelming majority of Wikipedia is not meaningless politics, but stated facts backed by decent sources.
This is true of good articles, but the overwhelming majority of Wikipedia tends to lack citations or, worse, cites sources that don't actually support the stated facts.
If an account in good standing adds a cited sentence the likelihood that anyone will actually go and check the source to confirm it supports the sentence is low. It's more likely that the edit will be reverted for other reasons.
Citogenesis is also a real problem, and wildly under-documented.
And most people who read Wikipedia do not take the time to examine all of the sources (if they're even able to - just cite a book if you want to make something up), read through the edit history, and get up to speed on the article-specific politics playing out on the talk page.
Still, it's better than everything else out there.
For a long time, traditional encyclopedias had a much better track record on topics related to politics and society, simply because their editor selection process largely eliminated single-issue crusading. You wouldn't be picked to lead a particular domain unless your academic track record made it clear that you're level-headed.
But I think that AI, just like your X friend anecdote, actually illustrates an interesting point: most of the time, when we consult some sort of an online reference, we're not doing anything important, so the accuracy is not critical. Quite often, we're just trying to validate our beliefs or win online arguments. An LLM that's 90% accurate but sounds 120% authoritative (and almost always willing to support your priors) is perfect for that.
> For a long time, traditional encyclopedias had a much better track record on topics related to politics and society, simply because their editor selection process largely eliminated single-issue crusading
That's a bit debatable. Traditional encyclopedias also had articles that were far from perfect, some of which had biases (not to mention there wasn't just one traditional encyclopedia. Different ones were of different quality). I think more research would be needed to figure out which is better.
Articles where almost everyone agrees on the facts are not interesting for discussing whether a particular encyclopedia is unbiased - it's precisely the contented topics where that distinction matters.
i'm not sure you're why you're being downvoted, relying on journalism is the weakest part of wikipedia, far and away, because it affects accuracy, which is what gell man amnesia is about, not bias. in comparison bias, in general, seems to happen regardless of sourcing.
He got done voted because that's an exaggerated claim. With no proof. Most of the articles on Wikipedia are just facts about nature, geography, history, etc and they don't need a newspaper as a source.
"Gell-Man" is an unfounded toy theory invented by an author without any research, using a colleague's name without permission to make it sound more authorative. It's hokum.
I got hit with a fraudulent chargeback (claim was the purchase was unauthorized and the person showed up in person to a class) and it was doubly bad because they paid via Link which means that Stripe actively verified them via 2FA.
Can someone explain to me why Stripe (or a competitor) doesn't offer a setting "refuse transactions for cards that have filed > x chargebacks with <acquirer> merchants this year"?
I don't know this is the reason, but if I were asked to build such a system, I'd be pretty worried that it constitutes a consumer report under the terms of the Fair Credit Reporting Act.
Certainly I wouldn't want the inevitable news drama about it. "I'm just a poor innocent grandma, I'm a trusting person when it comes to Facebook ads, and Stripe punished me for getting scammed by banning me from half the stores on the Internet!"
If your card is actually stolen then you should have the card number changed to prevent additional fraud and then the disputes would be against the old card number rather than the new one, right?
If your card is stolen you should, but not necessarily if you fall for a Facebook ad that ships you a pile of rocks or a paper photo of the product you thought you bought.
Isn't that exactly when you should have your card number changed? You gave your card info to a blatant fraudster. If they're willing to ship you a pile of rocks then there's a significant chance they're willing to use the card info you gave them to make fraudulent purchases.
Yeah, though this rule sounds a bit tricky. Like what if someone legitimately had their card abused.
The thing that gets me is that Stripe boasts about their machine learning radar rules etc etc, but somehow can't feed it actually valuable data.
Stripe support saw the emails from the customer boasting about defrauding me, they completely agreed that this is a clear case of friendly-fraud, but did nothing with this info.
Stripe can’t do anything per the way CCs work. Asking for that to change is a big ask. Asking my vendor to help me not do business with people who are likely to scam me is a smaller one.
100% agree. They cannot reverse the dispute. I already lost the money. I understand.
But Stripe is exactly in a position to at least use the evidence I provided (in this case, the evidence included the customer clearly admitting to friendly fraud), and feed it into their fraud-prevention system in some way. This way, lots of signals can help protect merchants from friendly fraudsters. So yes, I see it as a pretty small and legit ask from Stripe.
Do you imagine someone got a stolen credit card, made a linkedin with that name, used the card to attend a live class under the fake ID, or are you just doing the classic hacker news aaaaactually?
Comments like this have ruined this site. We all know that’s never happened once in history.
If you care about the quality of the site, consider the guidelines about not responding to a bad comment with a worse one and not griping about how HN has gotten terrible or turned into reddit or what there you. Downvote, flag, and move on to better discussion, and you'll spend a lot more time engaging and contributing to good discussions.
Contributing to good discussions is the highest leverage way to promote the quality of the site. Spending time in poor discussions is what makes it feel like HN has gone to crap.
The odds that someone meaningfully reflects on their behavior when you tell them they're what's wrong with HN are pretty slim. You can substantially increase those odds by being more neutral about it.
It's entirely possible HN is dying, HN won't last forever and dang has said he is very concerned about the direction he sees the commentary going. But it's also entirely possible that the quality waxes and wanes primarily due to internal factors and that political tensions, AI bots and slop submissions, etc are external factors straining HN that will eventually be resolved.
Either way, the "real HN" I find is still here if you look for it, and I have always seen people on here who complain about the quality of the site but when you go to their comment history (I have not looked at your comment history to be clear) you see they're constantly getting into arguments and architecting a miserable experience for themselves. One's experience on HN is largely what they make it.
allowlisting breaks once the agent has messaging tools. you can deny all outbound from the agent, but if it can post to teams or slack or email, link previews will fetch whatever URL the injection puts in. messaging is usually the first tool anyone adds to an enterprise agent so you end up with strict network controls that don't actually prevent anything.
I've built something similar to this internally and absolutely agree that timing is the real challenge.
Getting an AI to navigate a process is now easy. However, the difficult problem is having it understand:
1. How to screenshot or, even better, capture a GIF of exactly the right points in time
2. How to move the mouse and pause at the right points in time such that a human understands it
Isn’t the “solution” for Sentry that deploying it is such a pain in the ass that no one bothers to really do this? I haven’t checked in years but that always seemed like the real competitive blocker?
If you need less scale/features go for glitchtip. If you’re not going for k8s, the self-hosted docker-compose version of sentry works fine including proper releases and support by the sentry team etc. Just experimental newly introduced features can be a bit wonky.
They are doing much more than just throwing code over the fence. Also phone home telemetry is optional and there’s a switch for just errors mode. IMHO this really builds trust.
With regards to deployment complexity: well it’s built for handling high volumes of events. I’d reckon this is more a consequence of scaling the project rather than a coordinated plan to push people to their cloud offering.
If you do go for k8s or choose to deploy the stack yourself, you even get access to the full scale solution. But if you’re at that scale, you probably have someone hanging around who knows how to run your clickhouse setup. You still get the full sentry software and SDKs for free in that case. I think this is as fair as it gets with regards to the open source SaaS model.
This may very well be caused by my incompetence, but Sentry's docker-compose setup has never survived for more than a few months under my control. Something always destroys itself without an obvious reason sooner or later, and either refuses to start, or starts and doesn't really work. I tried updating it regularly, tried never updating it, getting the same treatment either way.
I did not intend to be critical of their work. They're doing OSS as best as they can and good for them. I am just saying that it's a different beast if Sentry is OSS vs a much simpler to operate OSS product. Licensing matters less when the operational cost acts as an inhibitor to adoption of your OSS offering.
True, opportunity cost is a factor, sorry if my reply sounded a bit brash. IMHO they are one of the few orgs who got this model right compared to lots of others who went the open core or support/consulting contract required OSS route.
Agreed. It was easier for me to rebuild parts of it for my own use than to self-host it. At my scale, a single DB works well as a datastore instead of Clickhouse/etc.
But then again I think this only prevents small players from "competing" by self-hosting, so the revenue loss there would be minimal either way. Large enterprises are too incompetent to even self-host a single self-contained binary, so for those the availability of source code and ease of hosting would make no difference, they would still use the SaaS.
> Isn’t the “solution” for Sentry that deploying it is such a pain in the ass that no one bothers to really do this?
That Sentry is a pain to deploy is not really intentional, it just happened over the years. However because it's a pain to deploy it also opens up a market for people that create managed deployments so I would say, that if anything, it made it worse. For self deployed Sentry you do not need to pay cent, the license explicitly allows it.
I'm personally on the fence how much of it is intentional... from the_mitsuhiko's side it probably isn't, but "the purpose of a system is what it does" and all.
Don't believe the salesmen, self hosting Sentry has been the most liberating feeling in a long while. Buy a cheap dedicated server with 64 gigs of RAM from Hetzner, run their install script and it's literally up and running. I'm processing volumes that would bankrupt me on their managed service without breaking a sweat.
I think this is a great idea with the wrong pricing model. Look at all the one off payment products that involve code, they're all dead. Just charge a lower but recurring price so I can be sure that you make enough money that you keep working on it. $20/month flat price to keep the license working and source available if you shut down. If people like it and want the barebones Sentry then charge them $40 a month to provide code and host it. Wishing you the best of luck.
ONCE projects do get occasional updates. I don’t use boilerplate projects much, so I can’t speak to them.
My model isn’t a subscription. Think about it like buying rice. You might buy it every week, but that doesn’t mean you’re “subscribed” to rice.
Even if I release a new major version, you’re free not to update. And if it’s a major version, it’s fair to expect it to be paid. After all, major updates usually bring significant improvements. For example, if you played the original DOOM, you had to pay for DOOM 2 too, even though they run on the same engine.
There are so many people buying rice that's not hard for rice producers to forecast how many thousands of tons of rice they have to grow each year. And not many people look at how much they spend for rice year long. For the single box, yes, they do notice. So it's not a subscription but it looks like it is, at least from the point of view of the seller.
Yes but I don't rely on rice I bought a year ago or DOOM as a core component of my business. Trying to work around a business model (subscription saas) requires that you understand what people are buying, and often, especially with you vendors, that is a financial alignment between the two.
If you bought DaVinci Resolve several years ago, you're still able to update to 20.<whatever> and use the same licence key.
Granted they're not interested in taking 225 quid off you for a software licence, they're interested in taking 22 grand off Netflix for a complete edit desk.
That's a fair point. In my case, though, I'm a solo dev without a hardware ecosystem, so major versions help sustain development without forcing subscriptions. What do you think about models like that for indie projects?
The problem with ONCE is that software is never finished. This is why most ONCE software that is still available today is charging a one off licensing fee + update fee (e.g. charge yearly for major updates or 10% of the one off fee per year). This is sustainable, but your model isn't. You will notice down the road in 2-4 years that it's no fun to work for free for users that expect an update because it requires patching or there are breaking changes.
That’s a fair concern, but I see it differently. Software can reach a point of maturity - not “dead” just done. That’s the whole philosophy behind ONCE: build something great, maintain it responsibly, and stop when it’s complete.
https://www.reaper.fm/ uses that pricing and has... let's say fanatical following. You pay once for version X and X+1 just in case you miss out on an update coming in a month. Then you pay again for a big upgrade.
Alert fatigue. Down Detector will show an outage with a service when the intermediate network is down. Companies have to triage alerts and once they’re validated they are posted on a status page. Some companies abuse this to hide their outages. Others delay in a reasonable manner.
I have considered building something to address this and even own honeststatuspage.com to eventually host it on. But it’s a complex problem without an obviously correct answer.
Yeah. Down Detector is more or less meaningless unless something massive has happened, and as you say it has terrible consequences for knock-on services.
It's not even just intermediate networks, it's sometimes direct coinnections. For example, a flood of people reporting an outage on mobile phone network X when the problem they are experiencing is not being able to call a loved one who is on phone network Y, which is the one that is down. This happened a little while back in the UK, leading the other phone providers to have to deny there was some broad outage (which is not an easy thing to reassure when there are so many MVNOs sharing network Y)
We're helping non technical users adopt AI by making usage visible. Power users become role models and people can upskill to gain value rather than lose their jobs to AI.
It's a complex problem because we need to balance trivial UX, with a slew of security and compliance requirements.
This is the first non founder developer role on the z0.ai codebase. We need someone who can own the full stack end to end: product surfaces people enjoy using, services that stay correct, and the taste to make both feel like one product.
We're funded and are live with our first customer already. We expect to onboard our next customer later this week and are actively ramping up sales efforts.
Check out the full job posting at https://www.z0.ai/careers/founding-engineer/ and send an email to join@z0.ai
reply