Charles Wright (@hackermath) header
Charles Wright (@hackermath) profile picture

Security and privacy research scientist / engineer. Formerly FUTO / Portland State / MIT-LL... PhD @ JHU.

DFW, TX
Joined July 2013
Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)Photo by Charles Wright (@hackermath)
Charles Wright retweeted
Legacy Media types are calling this Alex Karp interview a “crash-out” so that’s your first clue that he is actually saying something extremely insightful. He is articulating what real “AI safety” looks like in the enterprise. Not abstract alignment research or certification by a government-run DMV for AI. Real AI safety for businesses is the ability to control their own data, model weights, and compute — so a frontier lab can’t hoover up their proprietary knowledge and turn it into their next product. As Karp explains, technical customers want “control over their compute, their models, their data stack, and their alpha. They want to know they own the means of production, and it’s not being transferred to someone else.” Don’t think that can happen? Just look at Figma. According to The Information, Anthropic “blindsided” its then-business partner with the launch of Claude Design. Figma’s founder said Anthropic had not been “consistently honest” with them. Anthropic’s chief product officer had even served on Figma’s board until three days before the launch of Claude Design. Figma’s stock has fallen sharply this year while Anthropic’s valuation has surged. This isn’t an isolated example. Anthropic has launched Claude Science, Claude Security, Claude Legal, and of course Claude Code — each expanding into categories previously served by companies building on top of their models. The pattern is consistent: watch where value is being created, then move in directly. Dominate the model layer, then use that position to capture the most lucrative verticals. Dario has argued that open source models powerful enough to compete with Anthropic are “dangerous.” But dangerous to whom? Not to enterprises that want to retain control over their data and workflows. Dangerous to a business model that benefits from customers having few real alternatives at the model layer. As Karp exposes, true enterprise safety isn’t trusting that a lab’s future roadmap won’t include your business. It’s retaining the ability to choose — at the model layer — who gets to see and use your alpha.
Palantir CEO Alex Karp on what customers actually want, the real business of frontier labs, and the importance of open source models: “What the technical customers want is control over their compute, their models, their data stack, and their alpha. They want to know they own the means of production, and it's not being transferred to someone else.” "Who owns the data? Are the prompts secure? Is this being transferred to you?" "If it was so valuable, and I can make you a billion dollars, wouldn't I say I'll make you a billion dollars and I want 30%? Why are they charging for tokens if it's so valuable?"
647
1,836
285
13,308
2,152,988
Charles Wright retweeted
Replying to @jon_charb
The dystopian part isn't even privacy for age verification. There exist privacy-preserving proposals for that; Google and Apple are both building them. Google has a prototype. The dystopian part is: you likely can't make throwaway accounts to criticize your government. 1) The government can simply not give you an ID. 2) As we increasingly want proof that the person posting is a real human, we will limit how many accounts you can make, even privately, with a national ID card. 3) Anyone who gets access to your phone, e.g., at a police checkpoint, can probably use your ID to recover your burner accounts.
1
9
610
Charles Wright retweeted
I don’t think that academics should do ‘hot takes’ on matters of the day. Their opinions are rarely better informed than anyone else’s and cloaking them in scholarly garb cheapens the principle of objectivity. With respect to what is occurring in Britain today, in my opinion a fundamental error right now is to let the churn of the daily ‘news’ cycle drive your analysis. The legacy media, the government, and the police have all forfeited any claim to credibility; they lie routinely, by omission and commission, and they are actively shaping the narrative to protect a failing political order. Strong-arming victims’ families, suppressing footage, and spinning every incident as isolated ‘far-right thuggery’ or random criminality is not journalism or policing, let alone governing—it is damage limitation for a system that has lost control of the streets and the story. Instead, fix your gaze on the structural factors. Demography, geography, economics, and the hollowing-out of institutional legitimacy matter far more than whatever grainy mobile-phone clip is being waved at us this week. Britain has imported, at scale and with minimal integration, populations whose cultural distance from the native majority is large and, in important respects, growing rather than shrinking. Parallel societies, concentrated in particular towns and cities, now possess the critical mass to sustain sustained low-level conflict and, when conditions align, more organised violence. The state’s monopoly on force is visibly fraying; its willingness to use what remains of that monopoly is selective and therefore delegitimising. Trust in the police, courts, and political class is in the basement and still falling. Economic stagnation and housing pressure sharpen every grievance. These are not transient conditions; they are the terrain on which coming events will play out. On the Belfast attacks specifically: the operators are clearly more security-conscious than has been the case with the migrant hotel and other protests over the last couple of years—masked, disciplined about visuals, limiting the evidential trail. Some attribute this to institutional memory of the Troubles. That may be part of it. But I suspect the more immediate and probable vector is simple tactical diffusion from the modern Left and anarchist playbook. Black Bloc methods, the utility of anonymity, the selective application of violence, the media choreography—these have been field-tested and refined for years in Europe and North America. The manuals are not secret and the examples are legion: Marighella’s Minimanual of the Urban Guerrilla, the writings of the Red Army Faction, Alinsky’s organising principles stripped of the moralising, the operational literature of the Global Justice movement and Antifa networks. Remove the Marxist dialectical claptrap and you are left with cold, competent observations about how small, determined groups can punch above their weight against a larger but slower and more constrained opponent. Diffusion of those techniques was inevitable once the incentives aligned. You don't need an aged ex-IRA uncle to tell you how to do these things. The internet and a library card will do it. I am wary of firm day-to-day pronouncements precisely because reliable, on-the-ground reporting is so thin. I am not in Belfast, the journalistic desert in this country is real, nearly every dead-tree media and teevee pundit is a literal know nothing. What I will say with higher confidence on account of my reading of such conflicts elsewhere in the world is that certain escalatory dynamics are now highly probable: Police over-reaction that produces a martyr or martyrs, further radicalising elements on all sides. Targeted assassination of a judge, prominent politician, or influential voice. A spectacular, Christchurch-style mass killing when some individual or cell concludes that only dramatic, indiscriminate violence will break the equilibrium. Stabbings and gang rapes will continue at their grim baseline; they are already normalised enough that they barely shift the political dial. The deeper pattern is polarisation, erosion of restraint, and the slow emergence of organised ethnic and ideological blocs willing to use force to defend or advance their interests. All of that is in accordance with the rules of the game of identity politics, which were created by the *very same* people now most frantic about the perilous consequences of their own ideology. The centre is not holding because it has spent years delegitimising itself and disarming its natural supporters. Watch the structural trends—demographic momentum, institutional decay, the diffusion of effective small-group tactics, the collapse of shared reality—more than the latest headline. The news will keep lying. The underlying physics of the situation will not.
50
308
51
1,077
56,619
Charles Wright retweeted
This is an interesting change:
For the first time in 19 years, vulnerability exploitation is the #1 breach entry point. Not credentials. Not phishing. Software flaws, hit within hours of disclosure because AI compressed the attack timeline. The 2026 Verizon DBIR analyzed 22,000+ breaches across 145 countries. The data is clear: the window between disclosure and exploitation closed. Your traditional endpoints have runtime detection. Your AI endpoints don't. Every deployed model and agent is making decisions right now with no behavioral detections, no forensic trace, and no automated containment. That's the gap AI-EDR closes. For details, see: na2.hubs.ly/H060vtF0
4
10
57
15,643
Charles Wright retweeted
the art of technical writing is to appease both the p99 domain-expert and the curious p50 it's like a Pixar movie that speaks to child & parent
16
50
1
698
28,878
Nice write up from the Cloudflare team, but the post here is misleading. Patch faster is not the wrong answer, because most teams are patching on the order of weeks or months. You must patch faster than that right now. But I will agree that 2 hours is infeasible beyond the occasional emergency CVSS 10. And defense in depth is a part of the arsenal, now the whole story.
Cloudflare's security team spent the last few weeks testing Anthropic's Mythos against fifty of our own repositories. What we learned about offensive AI, why faster patching is the wrong reaction, and what the architecture around vulnerabilities has to look like next. cfl.re/49BRUqW
4
5
1
55
12,605
Charles Wright retweeted
I just learned the sad news that Peter Neumann has passed away. Peter Neumann shaped how a generation of security people learned to think about risk. As editor of RISKS Digest, he gave many of us coming up in the 1990s and early 2000s a steady education in the real-world consequences of computer failures. His work made the field more serious, more thoughtful, and more honest. He will be missed. I first met Peter when we both testified at the 1998 Senate Governmental Affairs Committee meeting on Government Security where the L0pht testified. The combination of Peter and the L0pht made the hearing more powerful even if us hackers stole the spotlight. Neumann and the L0pht made the same argument from two different directions. Neumann gave the institutional, systems-engineering view: the country was becoming dependent on brittle, interconnected systems that were never designed for security, reliability, or survivability. The L0pht gave the field evidence: here are the actual flaws, here is how attackers think, here is how cheaply and quickly these systems can fail in practice. Neumann supplied the credibility of a long-time researcher warning that this was not just “hackers breaking into things,” but a structural failure of technology markets, procurement, engineering discipline, and risk management. The L0pht supplied the proof that the warnings were not theoretical. Together, we made the hearing unusually powerful: the academic risk community and the hacker community were telling the Senate the same thing, in different languages, before the rest of the world had fully caught up.
9
32
5
90
18,417
I don't think we (academics) realize how vulnerable we are. I'm not sure how much longer tenure will last in an era when a) we've lost the public trust (for a lot of reasons, b) college/academia is a partisan political issue, and c) college-educated white-collar workers are losing their jobs to AI. We are in an incredibly privileged position relative to others, but that privilege also makes us far more precarious than most academics realize bc it makes us a target--and the threat is not just conservative politicians gunning for us but the much larger group of regular people who don't mind if we get put in our place or start losing our jobs like similar others or being asked to do tasks we don't want to do.
61
57
20
589
71,849
Exhausted zombie Pogi is not something you see very often
🗣️ La frustración de Pogacar tras ser segundo de nuevo en Roubaix por 'culpa' de Van Aert: "Me di cuenta que sería imposible al 99%" #ParisRoubaix
56
NEVER DOUBT
🇫🇷 #parisroubaix “This victory means everything to me." - Wout Van Aert Read more in our race report! 📷🗞️ teamvismaleaseabike.com/race…
28
Charles Wright retweeted
Yep, the Computer Security AIpocalypse is happening. Please be patient with the computer security professionals you know for the next few months, as they’ll likely be busy and stressed about preventing the collapse of civilization. teapawt.org/zooko/status/203878394…
11
34
12
326
67,672
Milan-Sanremo was awesome this year. * Pog showing that he's human after all * Pidcock showing that he's one of the top guys now * Van Aert showing that he's still got it * Pedersen signing up 5 minutes before the race and still beating literally everyone else
1
24
Charles Wright retweeted
When youthful energy directs itself toward productive paths it’s amazing what can be achieved. When directed toward the wrong paths it’s amazing how tons of hard work and enthusiasm can produce nothing of value. Trying hard and having a just goal isn’t enough. You actually have to be effective in order to make a difference.
37
27
5
350
22,423
A few years ago, I wasted way too much of my time and Eron's money trying to build something on top of Matrix. Should have built our own E2EE from scratch. Expensive lesson to learn.
Would it be fair at this point to declare the entire Matrix project to be a dead end? Looks like they've had tens of millions of dollars of funding with terrible results. The people involved seem incapable.
1
1
120
Someone builds a project management tool with Claude Code over a weekend. Ships it. Tweets "just replaced Jira." The app works. One user, happy path, localhost. Then two people edit the same record simultaneously, and the data is silently corrupted. They don't know what an optimistic lock is. They never needed to before. The prototype is maybe 1% of what makes software actually work. The other 99% is what you find after real users show up: race conditions, failed transactions, sessions expiring at the wrong moment, a payment webhook that fires twice and charges someone double. AI didn't cover any of that. It built exactly what you asked for. And the confidence is the worst part. "Just need to adjust a few things before we go live." The few things you need to adjust are the product. That's like laying a foundation and telling people you basically built the house. Vibe coding works. For personal tools, throwaway scripts, and prototypes you'll never put in front of paying users, it's genuinely fast and good enough. I use it. But there's a hard ceiling, and it shows up the moment the stakes get real. Agentic engineering is a different discipline. You're not prompting for code. You're decomposing problems, designing system boundaries, writing specs precise enough that the agent doesn't go sideways. You review everything it builds, because it will make mistakes that only look wrong if you know what correct looks like. You guide it. You catch what it misses. If you don't know what a distributed transaction is, the agent won't save you. It'll generate something broken with complete confidence, and you won't know until production. The hard part of software was never writing the first 200 lines. It never was.
174
253
47
1,983
202,258
I have to admit, I lol'd
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
1
2
170
Charles Wright retweeted
You may have noticed that when you ask an AI chatbot to evaluate ideas, it often starts with enthusiasm. And when you criticize its enthusiasm, it drops the enthusiasm. I asked Claude (Sonnet 4.6 extended) why it did this, and it had this reply—not news, but worth emphasizing.
11
26
7
130
20,576
When I was in high school they had us read something by Steinbeck. Was it The Grapes of Wrath, about courage and dignity? Was it the rousing and delightful Cannery Row? No. They made us read Steinbeck's The Pearl, a gloomy and horrible tale about a kid dying. F*ck you lady English teacher. Is it not okay for high school kids to read stuff which is uplifting? I know their brains aren't yet developed, but that, to me, is a really good reason to let them read stuff that will wire them to LIKE books. The Pearl made me never again want to read anything by Steinbeck. The Great Gatsby made me never want to read anything by Fitzgerald. I kept reading because I had been hooked on it as an 8 year old, but I learned pretty fast that every require reading in high school sucked donkey goober. Look - let the kids read Ivanhoe or Wind in the Willows or Penrod as high school kids. THEN, when they get older they may choose to dabble in the depressing crap you teachers had shoveled into your heads in college. But please let them learn to enjoy reading. I'm looking at crap like YOU Bridge to Terabithia. I've shown an image of a super fun book for teens.
I remember when I was in highschool I asked my English teacher: how come we only read books about sad losers who ended up more sad instead of stories where people have an arc and accomplish great things. She said we can't, we don't know how
179
48
25
788
58,646