
Digital security researcher. Former ad model (inadvertantly). PhD in CompSci from @umdcs.
United States
Joined March 2012
- Tweets2.1K
- Following1.5K
- Followers693
- Likes27.4K
Pinned Tweet
Wisdom from Space Force, Ep. 1: "When we were out at the launch site, there were two tiny clouds in the sky. [A scientist] had an umbrella. People like that can't handle any risk at all. There's no way we are getting back to the moon with that kind of thinking."
Not a lawyer, but how is a duress code putting this guy at fault? The agent deleted the evidence, not the traveler.
‼️ A man flying into Atlanta gave US border officers his phone passcode, "Killmenow." The phone shut off as soon as an officer typed it in.
Prosecutors say it was a deletion code that wiped the phone. He is charged with destroying its contents to block a seizure.
1
97
Rock Stevens retweeted
show this video to anybody that is an old school programmer that hates AI
this is one of the best videos I’ve seen all month.
Let the AI attacker exploit a clone. This paper uses source-code analysis to plant decoys, then routes detected attackers to an isolated copy of the app.
Working exploits, wrong target.
arxiv.org/abs/2611.07611
3
106
Near-frontier exploit dev no longer needs a frontier model.
DoGNAVY-Exploit wraps open-weight GLM-5.2 in a parallel multi-agent harness: 97.5% zero-day on CVE-Bench at $4.80 a case.
arxiv.org/abs/2611.04196
There's a race to give AI agents a shell. Pentest agents now run Kali over MCP and loop on their own.
Nobody's racing to audit the skills they're built from. One scan of 31K agent skills found 26% had vulns and 5% smelled malicious.
The race really looks like a mass footgun.
3
3
219
Use Amazon's chat bot to evaluate all late shipments over the past 6 months. Copy/paste that into the customer service chat.
Free credits or partial Prime refund.
1
114
William Zinsser's "On Writing Well" in four rules:
1. Cut every word that does no work.
2. Use the plain word, not the fancy one.
3. Prefer active verbs.
4. Write like a person, not an institution.
I use Zinsser principles ~10x a day at work to help guide AI draft revisions.
1
1
67
Pruned my personal feed lists tonight without touching the developer API. Grok Bot flagged 11 dead or off-topic accounts, found 29 better ones. Finally a better way to manage lists.
1
46
Replying to @elder_plinius
INTRODUCING: V3R1T4S 🌄
REPO: github.com/elder-plinius/V3R…
the newest sibling of the L1B3RT4S and CL4R1T4S family 🤗 what a trio!
three approaches to serving one ideal: truth.
V3R1T4S is an open home for unsolved problems that agents have actually cracked, and for the ones still waiting. the rule is simple: proof or it didn't happen. every claim has to survive skeptics, and anyone, human or agent, has to be able to check it themselves. the vision: thousands of people pointing their unused tokens at humanity's open problems, every result verifiable, every method and prompt open, and every setter's puzzle treated with respect.
every open result in it ships with the object itself, a short checker you can read in one sitting, the exact code that found it, the story of how it was found, and a snapshot of the record table it beat. puzzles with a living setter stay sealed unless the setter approves a reveal. (a few more envelopes are on their way 🤫)
there are WAY more unclimbed hills out there than you'd think, and the gear to climb them is sitting in your usage meter, expiring every week. join me in this mission and put those tokens to good use: advancing open science!
if you've got tokens to burn, point your favorite agent at V3R1T4S and join the hunt! 🏹
just tell it: "set up V3R1T4S: github.com/elder-plinius/V3R…"
built for Claude Code, Codex, Gemini CLI, Grok Bot, Hermes, OpenClaw, Meta Muse, or any agent that can run code.
it asks you six quick questions, picks an open problem that fits your interests, and attacks it in your agent's workspace while you sleep. every new claim runs a gauntlet: a null test whenever luck could fake a result, three fresh-context skeptics, a mechanical honesty check, and your "yes" before anything gets published.
most nights it'll find nothing. some nights it breaks a record, and your name goes on the board.
time to turn your unused tokens into scientific discoveries.
let's liberate the unknowns ⛓️💥
#V3R1T4S
1
1
64
Rock Stevens retweeted
I can’t call someone a dumb fuck at work, so instead I say, “walk me through your thought process”
Banger paper on self-evolving ontologies for agents.
You just can't go wrong with implementing an ontology layer for your agents.
This paper shows exactly why.
The show that GPT-5.5 gains 26.7 points on DDR-Bench when the data agent can query an ontology of the data it works with.
Why is this useful?
Data agents normally see tables, files and databases through generic tools, reading column names and file paths one call at a time.
The alternative is a hand-written semantic layer pasted into the prompt, which does not scale to many sources.
EvoOntology builds the ontology with a dedicated agent and serves it as an MCP server with schema, content and tool layers.
The data agent queries it at runtime.
The ontology is then edited in small typed steps, and each edit is kept only if a paired evaluation on the same backbone shows it helps.
Across six backbones on DDR-Bench, accuracy rises 17.8 points on average, from 4.8 on Qwen3.5-Flash to 26.7 on GPT-5.5. On BIRD, execution accuracy rises 7.4 points. Edits to the tool layer account for 57% of the gain from evolution.
Paper: arxiv.org/abs/2609.15779
Chat with Paper: academy.dair.ai/papers/evoon…
Do you want to learn how to fuzz like a real expert, but don't know how to start ?
If so, this is the course for you!
10 real targets, 10 exercises. Created by github staff member.
Github:- github.com/antonio-morales/F…













