(bre|m)aker. building @devicie, ctf @EmuExploit. securing agents, k8s, risc-v, etc. prev: gov, @ourANU
melbourne
Joined December 2019
- Tweets679
- Following922
- Followers394
- Likes1.6K
Tom Plant retweeted
I’m seeing a lot of Slibes (slop-vibe-slides) lately. AI can throw together a pretty-looking slide deck faster than you can blink. However, just like slop-text, slop-slides show a lack of respect for your audience. If it’s worth my time to read, it’s worth your time to edit.
Tom Plant retweeted
EMUCON 0x00 - "PILOT". Thank you again to our speakers, organisers and attendees. See you next year...?
Tom Plant retweeted
one of my engineers reached out to me about how they struggled to find satisfaction in their work in the age of AI. I asked them if I could publish our brief exchange, and they agreed. I know people are feeling similarly, so maybe they can relate. shubs.io/do-we-still-enjoy-s…
so you can pop calc, but can you crash it? bonus points if anyone golfs this
ty @sealldev & @UnpromptedAU for good vibes while testing this
Tom Plant retweeted
The BSidesCbr @sk8boardingdog CTF is LIVE and registrations are open!
1st: $2,500 + limited edition trophy + LEGO Donkey Kong Arcade
2nd: $1,250 + medal + LEGO Arcade Pinball
3rd: $750 + medal
Plus challenge prizes.
Register, rules & prizes:
2026.sk8.dog/
Tom Plant retweeted
When I created Raspbian for the Raspberry Pi, I was excited to help bring low-cost compute into the hands of a next generation of tinkerers and programmers.
However, I now despair that a Raspberry Pi can cost upwards of $300 because AI shortages are making personal computing unaffordable. Furthermore, bright young people are being discouraged from learning programming because they are being told it is now a commodity skill that AI will do.
I have to ask myself what it was all for. Is this the future we were promised? If so, I'm starting to regret my part in it.
Tom Plant retweeted
Latest preview build (29661) ships wesp.sys! The long-awaited initiative that should make it easier for EDRs to move capabilities to user mode and stay more resilient against kernel-level attacks.
Tom Plant retweeted
leaking Microsoft employee secrets through diagnostic data: careful.net/blog/powerlift/
Tom Plant retweeted
Of course that’s your contention.
You’re just getting into an RL data startup.
You just got finished reading some thread about how the harness is better and the scaffolding is what makes the difference.
You’re gonna be convinced of that til next month when you decide the model is what actually matters, and then you’re gonna be talking about pure capability and how harness doesn't mean much. That’s gonna last until you read more tweets and say the old model vs harness dichotomy is already obsolete and the RL-harness lifecycle with pure outcome rewards fail on long-horizon non-verifiable work because of credit assignment then talk about process supervision without knowing what it even means.
Then you're gonna read some technical blogpost explained to you by sol as ELI5 about how cheap reliable process-level signal+solving credit assignment at agent scale is what actually matters.
and in the end you'll realize that you've been following what everyone is coming up with no original thoughts while burning VC money because you can't even do B2B sales properly when you could've just done powerpoints made by LLMs and shilled it on linkedin
Tom Plant retweeted
If you like my reverse-engineering, deobfuscation or malware work, I'm looking for a job - I am a security researcher with a knack for deep tooling/capability development and I've also done blue-team/pentesting work.
DM me if you want my resume or if you want to ask questions!
Tom Plant retweeted
Things are getting weird (or interesting?) in the vulnerability research space. I published some of my personal thoughts on what we're seeing and what our broad strategy is here:
shubs.io/frontier-class-vuln…
Tom Plant retweeted
The launch of Stacked PRs is one of the biggest releases in @github history, touching almost every service we run
I wanted to share some details about how we built it and everything that went into it
Strap in folks, this is gonna be a long one 🧵
Tom Plant retweeted
Our innocent harness misconfig, their egregious misalignment
In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three different organizations.
Our post describes what happened, how it happened, and what we’re changing. We encourage other AI developers to perform similar reviews.
We conducted this review together with @Irregular, one of our evaluation partners, and thank them for the joint investigation and their collaboration on this post. This type of collaboration is increasingly critical to safe, rigorous evaluation of models, and we look forward to continuing to work together on security.
anthropic.com/news/investiga…
gg, he's cracked. writeup on this will be wild
I reported a pre auth RCE in Wordpress - it can be exploited with no preconditions, patch your instances!
slcyber.io/research-center/w…
Tom Plant retweeted
the most effort I've ever put into a CTF challenge: with a novel format that maximizes the human impact in scoring
minions in 16k
guest authored for @ProjectSEKAIctf, happening this weekend!