A founder who burned $100,000 on stolen API keys explains how a PHP bug and a 15-cent Chinese model turned hacking into a hobby, and the day-long fixes that protect you.
Posted
yesterday
Duration
Format
Interview
educational
Views
97.1K
949 likes
57 · 43
Big Idea
The argument in one line.
Cheap, fast open models made mass hacking economical, and because API keys are the most liquid thing an attacker can steal, the defense is limits, key pointers and an always-on agent watching your billing.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
You run a business or side project that calls paid AI APIs and you have no spending caps or alerts on those keys.
You ship apps with AI help, use a hosted backend like Supabase, and have never audited the security rules the model wrote.
You want a plain-English account of what happened in the Black Wednesday week and why it matters beyond the headlines.
You have old WordPress or PHP sites running somewhere and are not sure they have been patched.
SKIP IF…
You want a technical write-up of the PHP vulnerability itself. The conversation stays at the level of impact and fix, not exploit mechanics.
You need verified incident data. Most numbers (uptime, losses, discounts) are the guest's own estimates from conversation.
TL;DR
The full version, fast.
A critical PHP bug that gave attackers root access, plus a new ultra-cheap DeepSeek model, made automated hacking affordable for anyone. Attackers mostly target API keys at platforms like Railway and OpenCode, resell stolen Opus access in China at steep discounts, and labs distill the outputs into better models, which makes the next round cheaper still. US models refuse security work, so victims end up defending with the same open models. The guests recommend spending limits, 1Password service accounts, Have I Been Pwned, passkeys, Google or Apple sign-in, and an always-on agent that audits billing and logs every 30 minutes.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Cold open on Black Wednesday, then the guest frames the week as the worst incident level since CrowdStrike and a collapse in cloud uptime.
02:14 – 04:37
02 · The PHP Bug
A root-access flaw in PHP, which still underpins much of the web, forced an FBI and NSA warning. The fix is a quick upgrade, but old servers and WordPress blogs are exposed.
04:37 – 08:22
03 · How AI Fuels the Attacks
A cheap, fast DeepSeek model makes it economical to scan the entire internet for exploits, and it is easier to jailbreak than US models.
08:22 – 12:06
04 · The Stolen API Key Cycle
Attackers hit Railway and OpenCode for API keys, resell Opus access at deep discounts, and labs distill the outputs into stronger Chinese models.
12:06 – 16:52
05 · Protecting Your Business
Spending limits, the irony of US models refusing defensive work, trusted-access programs, and storing key pointers with 1Password service accounts.
16:52 – 21:24
06 · Protecting Yourself Online
Check Have I Been Pwned, audit authorized apps, use 1Password or Apple passwords, sign in with Google or Apple, adopt passkeys, and use an agent to delete dormant accounts.
21:24 – 25:31
07 · Building Secure Apps
New models write safer code than most people, so prompt for security, do not override backend rules, and run an always-on agent that audits billing and logs.
25:31 – 29:36
08 · US Should Work With China
Worries about the frontier going behind closed doors, and a hot take that US labs should teach Chinese labs to build guardrails, with the difficulties acknowledged.
Atomic Insights
Lines worth screenshotting.
Cloud providers used to advertise five nines of uptime, about two hours of downtime a year, and this year almost nobody has hit even two nines.
The fix for the PHP bug was a five-minute server upgrade, but the exposed servers were old ones nobody was maintaining.
Attackers are not after your username and password. They go straight for API keys because a key converts directly into cash.
An API key is the most liquid hackable asset: it can be resold as inference immediately, while a stolen login still needs more work to monetize.
A model at about 15 cents per million tokens is roughly 100x cheaper than Opus-class pricing, which is what makes scanning the whole internet economical.
Hacking got cheap because of price, not capability: models could hack 6 to 12 months ago but it cost too much or hit US guardrails.
Stolen Opus keys are resold on dozens of sites at around 90% off, and the output is then used to distill better Chinese models.
The loop feeds itself: stolen keys fund distillation, distillation gives better cheap models, and better cheap models make more hacks.
US models refused to help investigate an active attack because the request looked like malicious cyber work, so the victim had to use the open model that was attacking them.
Store a pointer to a key via a 1Password service account instead of the key itself. It does not stop a hack but it makes the damage easy to cut off.
Signing in with Google or Apple means a leaked password does not unlock your other sites, and their fraud models shut down odd logins fast.
Passkeys beat one-time codes because they require a physical device and your face or fingerprint.
Authorized-app lists on Twitter and Instagram are a quiet backdoor: one phished authorization can give someone access indefinitely.
A 200-dollar-a-month agent subscription running a 30-minute audit loop reportedly prevented at least three incidents and about $30,000 in losses.
Pointing an agent at your old accounts and telling it to delete the ones unused for six months is a security task that was not possible before.
Takeaway
Why API keys are the real target
WHAT TO LEARN
Cheap models made attacking everything affordable, so the practical defense is to assume keys will be targeted and make theft capped, cut off easily, and noticed within minutes.
02The PHP Bug
Cloud uptime collapsed this year, which means you should plan for outages and breaches as routine events rather than rare ones.
The PHP fix is a five-minute upgrade, so the risk sits with old servers and WordPress blogs nobody has touched in years.
03How AI Fuels the Attacks
Hacking got cheap because model prices fell by about 100x, not because models suddenly learned to hack.
04The Stolen API Key Cycle
Attackers skip passwords and personal data and go for API keys because a key converts straight into resellable inference.
Stolen keys fuel distillation of better Chinese models, which makes the next wave of attacks cheaper and keeps the loop turning.
05Protecting Your Business
Set hard spending limits on every AI API key. The guest's company lost around $100,000 to attackers running up a key.
Store pointers to keys with 1Password service accounts so any compromise can be cut off quickly.
US models may refuse defensive security work, so apply early to the trusted-access programs from Anthropic and OpenAI.
06Protecting Yourself Online
Run your email through Have I Been Pwned and reset anything that shows up in a breach.
Review the authorized-apps list on Twitter and Instagram, since one phished authorization gives lasting access.
Use passkeys, Google or Apple sign-in and a password manager with built-in two-factor codes for the strongest everyday protection.
Ask an agent to check your emails for breaches and delete accounts you have not used in six months.
07Building Secure Apps
Prompt your AI builder to make the app secure, and do not override the security rules on a hosted backend.
Run an always-on agent every 30 minutes over billing, logs and transactions so problems are fixed and reported while you sleep.
08US Should Work With China
The frontier may move behind closed doors, so expect the best models to arrive through restricted programs first.
Glossary
Terms worth knowing.
Black Wednesday
The name being used for the week a severe PHP bug, Node.js and GitLab issues, and a wave of API-key thefts landed together.
Five nines
99.999% uptime, which allows about five minutes of downtime per year in strict math and roughly two hours in the guest's rough framing.
Distillation
Training a smaller or newer model on the outputs of a stronger model so it copies much of its ability at lower cost.
Passkey
A login credential tied to your device and unlocked with Face ID, Touch ID or a PIN, so there is no password or code to steal.
Service account (1Password)
A 1Password feature that lets servers fetch secrets on demand, so the code stores a reference to an API key rather than the key itself.
Have I Been Pwned
A long-running free site that tells you whether your email or password appears in known data breaches.
Guardrails
Restrictions built into a model that make it refuse certain requests, such as helping attack systems.
Always-on agent
An AI agent that runs in a loop on a schedule, here checking billing, logs and transactions and alerting or fixing problems.
“No one has got even two nines of availability this year. So that's how bad it's gotten.”
Concrete, surprising number that needs no setup→ TikTok hook↗ Tweet quote
05:20
“Sixteen year old Billy Bob in whatever country can just basically ask DeepSeek V4 Flash to go look for these vulnerabilities.”
Makes the cost shift tangible in one sentence→ IG reel cold open↗ Tweet quote
56:46
“An API key is equivalent to cash. It's the most liquid form of hackable item that exists.”
One-line explanation of what attackers actually want→ newsletter pull-quote↗ Tweet quote
16:15
“We're stuck using swords to fight against swords, even though we have guns.”
Vivid analogy for the guardrail paradox→ TikTok hook↗ Tweet quote
23:16
“Have an always-on agent that just checks all your billing, logs and transactions every 30 minutes: is anything bad? If so, fix it and message me.”
Actionable and complete in one breath→ newsletter pull-quote↗ Tweet quote
Topic Map
Where the conversation goes.
02:14 – 04:37densePHP vulnerability and patching
04:37 – 08:22denseCheap models and automated exploit hunting
08:22 – 12:06denseStolen API keys and distillation loop
12:06 – 16:52denseBusiness protections and guardrail paradox
16:52 – 21:24densePersonal account security
21:24 – 25:31steadySecure app building and always-on agents
25:31 – 29:36steadyFrontier access and US-China guardrail cooperation
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
metaphor
Everyone is talking about Black Wednesday because the internet seems to be breaking in real time. And I'm going to be honest with you, this is the scariest week I've seen in AI. PHP, which runs about 70 % of the web, had a bug so bad the FBI and the NSA said patch your application or take it off.
line and at the same time hackers were stealing api keys and reselling them on chinese black markets this used to take an entire team of hackers now it takes one guy a cheap ai model and about 50 bucks so today i brought on anj nanda one of the best developers i know to break down everything that's going on okay anj my first question to you is what the hell is going on there's a lot going on i mean this is the craziest week in cyber for...
Ever since, you remember CrowdStrike going down like a year ago? Like this is probably the worst level of incidents that we've ever seen. You know, I think the way to quantify it is that before AI was a thing, all these cloud companies like AWS, Google, Microsoft, Azure, they were like touting themselves and having five nines of availability, which means 99 .999 % uptime.
That means they only go down for like two hours a year. For context, No one has got even two nines of availability this year.
So that's how bad it's gotten. It's actually gotten three orders of magnitude worse than it used to ever be. There were two main things that happened.
One is that we found a bunch of bugs in some core technologies that power over 90 % of the internet, which is in PHP and in Node .js, as well as GitLab, which hosts the code for almost all enterprises. And it's the second largest Git provider other than GitHub. And then we saw a whole host of attacks that I've actually seen up close and personal firsthand, where we saw these cyber criminals essentially try to hack infrastructure providers like railway and open code to steal ai credits from them and they wanted to steal these credits to to sell them so we can talk more about that later but that's basically what happened okay so this is let's let's break this down so can you talk about like what the specific bug is or bugs are why it's happening now and like what models people are using to like exploit these vulnerabilities yeah that's a great question so php basically runs most of the internet and it might not be even even if you're coding your website in PHP.
No one really uses that anymore. For context, the last big website that was made in PHP was Facebook, and that was started over 20 years ago. Ever since then, people stopped using it, but it still powers a lot of the internet because a lot of the core services that power the newer technologies is built on PHP.
So a lot of coding is just built on the old stuff, and PHP is one of those foundational level things. And the bug was essentially that hacker, if they could exploit the bug, was able to get root access to all the files on your server.
So it was, like, really, really bad. So they could, like, get access to everything. So they literally could make changes to an entire app.
They could make changes. They could read it. They could read everything.
They could get all your back -end data, which is, like, all your secure data that's not supposed to be publicly available. You know, most people were, like, prevented from this being an issue because, you know, they are using another service provider that automatically upgraded. But some of the older websites are, like, the ones that really suffered from here.
Like, if you ever had a WordPress blog, you know, got to check. out because it's probably probably gone at this point yeah that is concerning right uh millions of people have used uh wordpress yeah okay okay so okay so there's this vulnerability and i guess we don't need to dive too deep into like actually how that came to be i guess i guess the next question would be i think that nsa or in the fbi basically said you need to patch this so there was this vulnerability and it was up to the people who built these applications on php to like patch them like what is patching that me like can you describe what that actually means yeah the actual fix is really simple you just upgrade your server so you take it down for like five minutes and you upgrade it like it's really simple but you know these are a bunch of like really old servers most likely that's why the nsa and fbi released this security warning and they were like if you can't take this if you can't upgrade your server like right now just take it down for a few hours because that's how bad the exploit was remember this export gives any attacker full access to everything on your server so how does ai play a role how does ai play a role in this like is it
That's a great question. Okay, yeah, go ahead. So 15 days of these attacks.
we got a new model from our favorite people across the pond, DeepSeek. So they created DeepSeek version 4 .1 Flash, which is an extremely cheap model, and it's really good at coding, much better than anything in the past, and it's really fast. So this combination of attributes for this model made it so that now it makes economical sense for attackers to basically go and tell their model, hey, why don't you...
like go find every website on the internet and try to hack it basically okay so so so basically like It is cheaper and faster than ever to go look for these exploits where in the past it would require a human. And that's also a niche skill, right?
Like not everyone would know how to go look for exploits. And so that's why it was done by like, you know, at least in my imagination by like hooded criminals that are, you know, in some warehouse somewhere. But now 16 year old Billy Bob in whatever country can just basically ask deep seek V4 flash to go look for these vulnerabilities.
And now it's now. It's like exponential. I'm seeing charts.
I'm seeing charts of people. I forget what they're called, but like showing how it's like exponentially worse, like over the last three days, like the amount of these vulnerabilities found are going up exponentially. I think it's all based on like the deep seek models, because even though AI has gotten good enough to hack maybe like six months ago or even like 12 months ago, it was just not economical or it was not allowed because of the guardrails of the U .S.
models so if you know anthropic they really care about security guardrails that kind of stuff so they made it really hard for you to use their models in a way that they didn't want you to use them but now because deep seek version 4 .1 flash came out and it's super cheap it's like 15 cents per million tokens which is I think two orders of magnitude cheaper than sonnet and maybe like four orders of magnitude cheaper than you know opus it's just 15 cents per million tokens out yeah yeah like opus is what like 50 dollars yeah 35 something like that yeah yeah so it's like two orders like 100 x you know cheaper and also they're really easy to jailbreak.
So I actually don't think that these Chinese labs are evil. They're not like, ooh, we're going to release this model to make it so that we will have overthrown US. I think China is fiercely competitive.
If you ever saw Travis Kalanick from Uber talk about how China was competitive, you've seen that, right? What did he say about China? He was just like, you know, they just...
ruthlessly copy everything. So China is at the point where their tech industry is looking at America and they're like, wow, we're lagging behind. We just want to copy everything.
And they're not doing it with mal -intent, but they don't have the same level of care about security. So I actually spoke to the CEO of Xiaomi. which is one of the biggest publicly traded technology companies in China, which is, you know, they're kind of like, let's say, the Google of China.
And he said that he wants US... Model companies to help them figure out security and like tell them what to do because they don't know where to even start Putting guardrails on these models and the US model companies don't want to talk to the Chinese model companies because they're basically stealing their Models and we can get into that.
So this is actually biggest reason why these hacks are happening because it's a vicious loop So for example, these hackers are exploiters in China will go hack Railway or OpenCode. And they won't even care about the user credentials.
Some of these hacks, they'd have access to your whole bank account or Google. Maybe not bank account, but your Google information and your personal data. They'll ignore that, and they'll just go to your API keys.
Because these hackers can essentially steal the API keys from companies like Railway and OpenCode, like developers who are developing on those platforms. Then they can get inference to the cloud models for really, really cheap. because China is restricted.
Remember, China doesn't have access to anthropic models. It's actually pretty hard for them to get access to it. So then these hackers take those models, and then they sell it to other people in China through their service.
So there's actually dozens of these websites that you can find online where I'm not going to tell you which ones, but you can buy. Opus at like 90 % discount to normal Opus pricing because people are literally stealing Opus API keys and like reselling these tokens. And then the labs just buy these inference from these providers and then they distill their own models like deep seek on the output of the cloud models, and then those models are then used to hack.
So then the hacks become even cheaper because the models become better, they become smarter, and DeepSeq is trying to race to the bottom in terms of price. And so it's like this vicious cycle where API keys get stolen, you have distillation, then you get better Chinese model, and then the hacks get cheaper, so more API keys get stolen, so there's more distillation, so you get even better Chinese model.
That's crazy. That's crazy. So this whole situation, is literally making the Chinese models better.
Yes, yes. The fact that the US is not doing anything or like is so restrictive to China is actually making the Chinese models better. It's like, you know, Jensen Huang said this on a podcast.
He was like, because we cut off GPU supply to China, they developed their own. chips to run their own models. So like all these Chinese models run their own chips because, you know, how do you make diamonds?
Pressure. Pressure creates diamonds. China is just in the pressure zone right now.
So they have to it's like do or die for them. Right. And they're raising billions of dollars.
So it's it's it's becoming even more crucial that they're competitive. So, yeah, this is not looking. This is very, this is a very interesting situation.
So I guess let me summarize what we've talked about so far. And then I kind of want to move to more practical topics like tactically, what do individuals do and what should businesses do as a result of this? But just to kind of summarize, so we have this terrible bug.
that was exposed in PHP and some other technologies. And, you know, people aren't necessarily using PHP as much as they are using other services that have been using PHP for a very long time. So it's a very foundational layer of kind of the infrastructure stack that we use for the Internet.
And it's starting to feel flimsy. And people are getting worried. And now, because AI models like DeepSeek 4 .1 Flash is getting faster and cheaper, people are using it to look for vulnerabilities.
And it's creating this feedback loop whereby as they are using these vulnerabilities to steal API keys, then there are these sites which basically have like thousands of API keys. And then they just drain them all. And you're able to use these at a discount, which allows them to gather more data, China to gather more data from frontier labs, which then they can use to make their models better.
So we have this vicious loop. And obviously. That sounds scary.
That sounds scary. And so now let's think about what we can actually do to prevent bad things to happen. So I guess to start, can you just outline some of the bad things that could, in theory, happen to me or my business because of this?
So if you're running a business and you use AI, which is probably most of you watching this, and you use API keys and you don't have limits on it, theoretically, a hacker could... get your API key and just spend it up. We've had that happen to vibe code, and you've seen that firsthand, right?
We've lost, I think, you know, luckily we've never been hacked and no user data was lost, but like we burned like, you know, $100 ,000 because of these exploiters who basically were able to find a bunch of free credits and then just like run up our API key bill, which is probably the biggest thing that can happen is like they just want your API keys.
That's it. And it's actually funny. Because, yeah, go ahead.
No, no, I think, yeah, I mean, I think it's easier said than done. Like at a personal level, or if you run a smaller business, you can set API key limits. But obviously, if you're a company that has thousands of users, that you could spend hundreds of thousands of dollars a month on API keys, it's probably harder to set limits.
So I guess, yeah, like what do you recommend doing in regards to like API keys getting drained? I think the most ironic thing is that we can't even use the US models to protect us from this incident. So OpenCode was getting hacked, and they were getting their API keys stolen.
And they were trying to use GPT -6 Astra being like, hey, someone's hacking us. Can you look into what's going on? And GPT -6 Astra was like, cybersecurity warning.
We're not allowed to look into this because this might be considered malicious. Oh, that's interesting. OK, so.
Oh, because, because, okay, okay. This is really interesting. So like, I remember reading about this where like, you know, when Fable 5 point, God, I'm just really struggling with Fable 5 came out.
Yeah. They weren't allowing anyone to do any cyber requests, right? That's how they were famously rerouting you silently to like a worse Claude model, like Claude Opus 4 .7 or whatever at the time.
And so. Because the way that you would get it to do nefarious things is trick it into saying that you're like stress testing your infrastructure. Exactly.
Yeah. But if something actually goes wrong, you can't even get Claude models to do anything because they're like, oh, no, they're trying to pwn us. They're trying to like jailbreak the model to get us to do bad things.
And so now, like, yeah, what did OpenCode have to do then? So they had to use the same open source models that were used to hack them. That's crazy.
So the most ironic thing is we have to use DeepSeek. We 4 .1 Flash to prevent us from DeepSeek trying to hack us. Well, I mean, maybe you could use like Kimmy K3 or maybe I don't even know what the stronger.
the stronger models, open models, but that sounds really scary. And I guess, is there some sort of program where OpenAI and Anthropic could go to trustworthy companies to give them access to these better models for these tasks or what? Yeah, OpenAI has a daybreak program.
That's what they call them. And then I think... Anthropic calls it like cyber protection program or something.
They have like cyber protection programs where you can go apply and then they like allow your account to access models that it's the same model, but like it has less guardrails. So for example, we applied to Anthropic's program, but honestly, these programs are. probably have a really long backlog right now.
And we haven't gotten access yet. And so it takes some time to figure this out. So actually, we're having to use the open source models to prevent us.
You know, we, it's like the U .S. invented, like, guns, and we're fighting against people who have swords, but we're not allowed to use the guns in battle because our leaders are like, hey, these guns are too powerful. We need to slow down the development of guns.
And so we're stuck using swords to fight against swords, even though we have guns. That's kind of like that. But for somehow they're stealing our guns and learning how to make better swords.
Yeah. Somehow. Something like that.
Sure, sure, sure, sure. But yeah, you can apply to them. I'm sure they'll figure out how to get more companies onto these programs now that this has become a big issue.
I think another big tip for businesses is to use the one password service accounts, which basically have pointers towards API keys. So then you wouldn't store an API key openly in your... server, you would just store a pointer to it.
And so it doesn't prevent the hacks, but it makes it easy to restrict them. in the future essentially so i recommend using one password okay and so it's not just api keys that are at risk right like also your data right your your your usernames and passwords and emails and other data has already been compromised and sold overseas i'm told already i guess a how do you know if you've been compromised if there is a way to know and then b like what should i do just as a user of the internet to make sure that my data is safe yeah i think number one go to have i been important .com so you can go and put your email in there and your password and you can see if it's in any of these like dark web public data sets the hackers usually the ones who hack you don't really care about your data they just sell them to other people who then do things with that data But if you go to haveibeenpwned .com, you can put in your emails and your passwords and you can see what's going on.
And it's secure? It's secure. It's been around for like 20 years.
I think created by some pretty great security engineers. So I do recommend it. You can put the official link.
But yeah, it's been around for a while. And I've used it many times and I've seen when I've been pwned and that kind of stuff. And I think 1Password actually sponsors it.
Okay, so it's a legit thing. It's a legit site. It's haveibeenpwned .com.
You might understand why I would think it's not secure. It's a funny name. Yeah, yeah.
Okay, so you can go to this. And obviously, if you've been pwned, then you should probably reset your passwords and such. Another thing, I got my Twitter hacked, I think, six months ago.
I think you remember that. And that was because I had a authorized apps that should not have been authorized. So this is another thing on Twitter.
Highly recommend looking into it, especially if you ever click links in an email like you may have and somehow they got you to like authorize your Twitter account. You may have someone who just has access at any time to make changes to your account. So the first thing you need to do is like go to authorized accounts on Twitter, Instagram, things like that.
So no one has access to your Twitter. I saw a guy and he's like an artist who got. hacked yesterday but regardless you should look at that and then also you should make sure that you have like two -factor authentication on like do you have any recommendations on how to do that or like what that's like should you use the like Google Authenticator app or whatever I recommend using 1Password I think if you have your passwords in there or and your two -factor codes that's great if you use iPhones and Apple devices you can also use Apple passwords I think that's also awesome.
But yeah, you should idly sign into any services using your Google and Apple account as well because they have the best security. So even if somehow someone gets your password and then they figure out your two -factor authentication, but it's coming from China, they're going to shut down your account, restrict it. They know the consumer patterns, and they have these really big machine learning models basically running 24 -7 trying to figure out, hey, is any of these accounts wonderful?
is like do I see any like odd behavior and they like instantly shut your account down and reach out to you sometimes even for like ID verification which is not ideal but like you know at this point would you rather lose your data or have Google have a little bit more information about you I mean they have everything already so I think I'd rather trust Google than like any random app developer like think about it right like if you go log into an account that's like a new app and you put in a username and password and it's the one you always use, if that gets leaked, then they can log into any other sites that also share that same password.
But if you just log in with Google, now you're trusting Google for that authentication for the password, and Google's probably not going to get hacked. So I recommend signing in with Google and Apple. Also, other than two -factor authentication, use passkeys.
Passkeys are even more secure than two -factor authentication because you need either a physical device to authenticate. to authenticate yourself, like Dutch ID or Face ID, which is even better than just an authentication code. And then you can actually use agents to go scour the internet.
So if you have Grokbot or ChatGPT dots, you just be like, hey, these are my emails. See if I've been pwned. And go to websites that I haven't used in six months and delete all of them.
And they'll just log into all these websites and delete all your data, which has never been possible before. I think that's one tip that I would really recommend doing if you haven't used these tools, that's the perfect use case.
Use agents to find these things and delete accounts that you don't use. I think that makes sense. Very interesting.
OK, so I want to talk about building apps. So obviously, you have a lot of non -technical people now building apps, whether that's mobile apps, web apps, internal tools for your company. I guess, what do you think the future of this is?
What are the precautions people need to take when building apps to avoid these vulnerabilities? this not that big of an issue because most people aren't building apps in PHP. Talk to me a little bit about that.
Yeah, I think the new models are probably better at security than most humans, right? So they're using newer technologies. They're not using PHP.
They know what to do to protect you against it. Just, you know, don't, I think a lot of people who've been phoned. I see them using Superbase and then overriding their security rules, which is a very common thing for app developers to do.
I just say that just prompt your AI, be like, hey, make my app secure, and it's probably going to do a better job than any human could do. They're better at us coding. And then another thing I recommend is having an always -on agent.
dot or grokbot or whatever you like even cloud code that's just running in a loop and it just checks all your billing infrastructure because that's the main exploit like these people like they've hacked like open code and they don't care about your username they just care about the api keys so that because that directly translates into cache if they just get your like username and password now they have to go find which other accounts like are you using the same user like it's a lot more work versus like an api key is like equivalent to cache it's basically the most liquid form of hackable item that exists.
It's like the most liquid thing. So, you know, have a always on agent that like looks at all your transactions and looks at all your data and like looks at all your logging and like just every 30 minutes is like, is anything bad? If something's bad, fix it and send me a message.
That's it. We do that. It's prevented us from like at least three.
of these issues and like 30 000 in losses at the minimum and it pays itself back like just 200 cloud subscription and you're good to go you know yeah yeah yeah and this isn't a new thing right like people have been dealing with hacks all year i mean for the last like six to 12 months they're just it's getting more accessible and i think yeah i guess that brings me to something i was thinking about yesterday i tweeted what do you think the last frontier model released to the public will be and i was thinking about that and i actually got really sad because um i guess AI for me has always been a thing where you could exist at the frontier just being a citizen, right?
You could just, ChatGPT was the best model that came out. I was using the best model when Sonnet 3 .5 came out. You know, the individuals were actually ahead of the government.
They were ahead of companies because they were first to adopt and it was the best model on the market. You know, and then I think Fable. was kind of the first time where I felt like, oh, wow, this is like the frontier is going to be hidden behind closed doors.
And now I just I can't help but get that feeling that that's only going to get worse as especially with the distillations, you know, putting these models out in the world makes these models that can do hacking much better. it seems to me that the models may be behind some closed -door thing, and that's kind of sad to me.
What do you think about that? Do you think we may be at the end of the publicly accessible frontier, at least in the near term? I mean, I hope not.
I feel like that would just make this process worse, because if we reach that point, it means that the models have gotten so good where, you know, the moment... And it's just a matter of time until they distill these restricted models, right? And the moment they do that, what then?
Then it's like OpenCode had to use open source models to protect themselves. We need to equip our country and our developers with the better models, I feel like. And if it's restricted, then only some companies will get access to it.
So I'm scared. I think it's like... You're racing against the clock, but then you're trying to add these restrictions.
I have a hot take here. I really think that the U .S. model companies should really work with the Chinese model companies and teach them how to build safeguards and guardrails into their models.
So even if these Chinese model companies get better, and they will, it's just a matter of time. They are already good enough. for these exploits, but they will get better.
But if they know how to do the guardrails as good as the U .S. model companies, then maybe these hacks will go down because the hackers won't be able to use the Chinese models to do their dirty work. So I think there should be more collaboration across the two major AI powers, which is contrary to what I feel like most people are advising.
It's easier said than done, though. You know what I mean? Yeah, it is.
Because not only do you need to coordinate with what some people consider an adversary, you also need to compete at the same time with companies inside your own country.
And you don't want to give away your best secrets because, you know, these are private companies at the end of the day. And then China could always defect too. Like they could take the knowledge from them and then remove the guardrails later because it sounds like they're literally benefiting from there not being guardrails, at least right now.
And so maybe there's some sort of law passed that requires any open models that are used to have some sort of certification that there are guardrails on. I don't know how this is going to work.
But yeah, it's easier said than done. Yeah, no, for sure. I mean, you're right.
Like, that's the risk, right? But yeah, I'm in the camp of... So what do you think?
You think we should just restrict China from using all these models? No, I have no answer. I have no answer.
I was just saying it's easier said than done. I think just because it's hard doesn't mean you shouldn't do it. If you were Dario, what would you do right now?
Dude, I can't be Dario. Dario, he thinks that he thinks it's such a different way than most CEOs. He thinks in essays.
And I don't think like that. I don't know. To me, you know, maybe I just think simply like I'm just scared that I won't get access to Fable 5 .5 or that it'll never be released.
I just want a better model. I just want to use the best models and make cool stuff, man. Yeah, this is interesting.
This is an interesting situation. Okay, so we have covered the bugs. We have covered the attacks.
We've covered the fact that it's cheaper and faster and anyone can kind of participate on it. Are we missing anything? No, I think that's pretty much it.
We've covered what you should do as a company, in a business, and also as an individual. And I think it's like very easy steps that you can take, like you can do today. Like we did this at my company and it took like a day.
And I'm sure you've done it as well. And it's like it doesn't take that long to protect against these things. But like I used to be the person who was like, oh, yeah, I'm never going to get hacked.
I'm a tech guy. But now I'm like my mentality is completely shifted. It's like it's no longer about like just knowing.
like a little bit, you got to have your agents protecting you 24 -7. So like as long as I'm sleeping at night and I got this agent running, you know, making sure everything's okay, that's like my... my safety net which is interesting it's like before we didn't need that now we need these like always -on agents as the only way is like it's like sort of the last line of defense now right i think this is a good emergency podcast yeah where can people find you yeah i just started making youtube videos you can search my full name on shananda on youtube feel free to subscribe to me and hopefully me and rally can do this again but yeah we we'll do a positive topic next time we'll talk about fun tech maybe we'll do like a review on opus or a fable 5 .5 when it gets released in uh 2028 uh after they implement guardrails all right uh thanks for for coming on i hope you guys gathered a lot of useful info from this just do the things make sure you're secure it's getting scary out there and don't search for vulnerabilities and hack people's api keys all right see you guys
The Hook
The bait, then the rug-pull.
The host opens by calling it the scariest week he has seen in AI: a PHP bug the FBI and NSA told people to patch or take offline, while hackers resold stolen API keys on Chinese black markets. Then a guest who lost real money to this walks through why it happened and what to do.
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
Riley Brown argues the loudest AI story is backwards: the people with the most agency are using it to delete slop, from TV bloatware to app upsells to bloated SaaS.
A 20-item weekly rundown of AI agent news: Claude Dashboards and Motion, GrokBot routing to Opus 5.5, ChatGPT's inline mini-apps, Nano Banana 2.1, and Adobe's open-source clones.
Five days into Claude Opus 5.5, one builder rebuilt Notion, MyMind, and Typefully himself, then laid out what OpenAI is rumored to answer with at Dev Day.
A two-hour, filmed-in-one-day walkthrough of everything Codex can do with GPT-6 Astra, from one-prompt iOS apps to an agent that reads your Slack and tells you what to work on.
00:27
00:44
01:24
01:31
02:25
02:38
02:59
03:37
03:46
04:15
04:57
05:18
05:41
05:58
06:28
06:49
07:05
07:25
07:57
08:09
08:53
09:21
09:58
10:10
10:38
10:54
11:20
11:31
11:55
12:52
13:12
13:30
13:45
14:14
14:36
14:55
15:20
15:43
15:59
16:20
17:05
17:24
17:55
18:18
18:34
19:10
19:25
19:47
20:10
20:21
20:48
21:49
22:00
22:19
22:38
23:16
23:36
23:46
24:03
24:39
24:58
25:42
26:14
26:22
26:44
27:11
27:33
27:46
28:21
28:40
28:52
29:26
28:05
27:34
20:50
21:08
Video of the Day22:27
1:54:47