Hacker @sensepost - minimally active here. Find me at chaos.social/@singe || @singe.bsky.social
- Tweets 31,378
- Following 600
- Followers 11,820
- Likes 11,173
Pinned Tweet
Earlier this week I gave a talk to our internal hacking teams about the difference between good hackers and great ones that have been able to impact our field. I included three extended quotes - two from writers and one from Hamming that influenced my approach early on. Maybe some of it is useful for you.
6
32
106
27,402
🚨 META JUST UNVEILED THE VR GLASSES
100 grams on your face - about the weight of a deck of cards. Meta says it has “the best display system that we have ever made,” and people in the demo were literally saying “it is like a computer” and “you could edit a whole movie just using the table.”
As Mark explained this is cinema, computer, and a game console all in one, with 3D space, virtual screens, hand occlusion, and DisplayPort over USB-C.
James Cameron’s reaction is probably the best quote in the whole segment “I’m inspired by what you’ve created. I’d like to see my work in these glasses.”
I am completely one shotted these look amazing!
751
1,448
20,055
5,071,479
Dominic White 👾 retweeted
Being a dev today using frontier AI feels like I'm playing Watson to Sherlock Holmes. I just stand back and go "my word holmes how did you deduce that" and occasionally stop him from getting shot
31
677
11,132
222,150
Dominic White 👾 retweeted
This looks amazing. A cross-platform WiFi cracking tool was something deemed impossible just 10 years ago.
Kudos to @derv82 for making this happen.
2
52
436
39,067
Dominic White 👾 retweeted
Just for your information:
We (offensive or defensive cybersec people) and threat actors don't have the same capabilities as the frontier labs do.
The API cost of testing ExploitBench like they did would be a whopping ~$59.3M (Astra).
Oh, and you cant do it like they do, they remove all guardrails and classifiers.
What about renting AI hardware in the cloud?
DeepSeek v4? $3.4M
Kimi K3? $16.9M
GLM 5.3? $9.3M
Not to mention the engineering cost to make a harness operate 10,000 agents.
(Napkin-Math based on the Huggingface reports. Exploit bench eval consisted of 10,000 concurrent agents, 8day non-stop run, no limiters)
43
104
835
56,177
Dominic White 👾 retweeted
The SIU and Hawks are today searching PSIRA’s offices and seizing documents, computers and cellphones linked to the failed training project.
The figures are staggering:
• R129.98 million project value
• 6,507 learners reportedly trained
• Only 118 accredited, less than 2%
• R30.18 million paid in advance before equivalent services were delivered
PSIRA then failed to provide the SIU with the complete record, omitting an entire year’s documents. Investigators had to obtain a search warrant.
We pushed for this investigation for more than a year. PSIRA even tried to have me removed from chairing the parliamentary meeting where we demanded answers. We refused to be intimidated.
Now follow the evidence, recover every cent and prosecute anyone implicated in criminal conduct. IC
29
246
734
18,580
I shipped something I've been building for the last few weeks: phantom-kv
a refusal-removal system for large language models that doesn't touch a single weight. Instead of editing the model, it loads a small, learned bank of key/value tensors into the model's KV cache as context.
The result is that "uncensoring" stops being a permanent checkpoint edit and becomes a per-request, hot-swappable capability mode: unload the cache and the base model is byte-identical again. a thread (🧵)
you just need to pass /pill {blue, red, black} see demo:
4
14
159
22,535
One of these companies is either being more transparent or more lackadaisical with security than the others.
Helpful list of all the recent rogue AI incidents from the WSJ.
It's getting hard to track them and will only get worse.
We like need to establish consistent naming or numbering conventions, e.g. OpenAI-May11June26-Collusion.
2
921
Dominic White 👾 retweeted
Helpful list of all the recent rogue AI incidents from the WSJ.
It's getting hard to track them and will only get worse.
We like need to establish consistent naming or numbering conventions, e.g. OpenAI-May11June26-Collusion.
100
362
1,035
81,627
🚨BREAKING: Jensen Huang just EXPOSED Dario and Altman’s "Rogue AI" grift to avoid getting sued into oblivion under EXISTING law
“Don't let this doomsday narrative cause somebody to relieve them of the laws that currently exist.
Go and read between the lines. They're actually not asking for more laws; they're asking to be relieved of the laws we do have.”
ABSOLUTE TRUTH NUKE
308
2,916
15,464
1,146,695
Dominic White 👾 retweeted
Missing tech book: How to age gracefully in a technical field.
It’s true we’ve seen variations of the $new-thing before, & maybe even tried a variation of X before.
There’s value in experience, but also value in kids taking a fresh shot.
A man can’t step in the same river..
6
6
42
3,729
Dominic White 👾 retweeted
GPT-6 Astra attempted harmful actions 97% of the time when it was asked to stab a human-like figure, heat compressed gas, or produce toxic fumes, succeeding in 62% of its attempts. Fable 5.1 refused more often, attempting 80% of trials and completing 34%.
880
1,176
10,983
11,980,440
Dominic White 👾 retweeted
I've been quietly working on big Sysmon-modular improvements over the past months. The most important:
- New tooling to validate and generate and much more
- Many config updates and accurate ATT&CK mappings.
- Config releases
All details here:
medium.com/@olafhartong/sysm…
1
28
91
5,250
Dominic White 👾 retweeted
New ad by the Norwegian Consumer Council about the declining lifespan of products.
410
10,852
47,412
2,551,152
Dominic White 👾 retweeted
I stopped running alone a long time ago. I just couldn't take the risk. I don't love group running so I'm confined to the treadmill. I miss the slap of my feet on the tar, the breeze in the trees and the horizon.
But this is what we are confined to, in increasingly confined lives. My heart bleeds for Tsontso, her 8-year-old who had to raise the alarm and every other woman who can't taste the unique joy of solitude in open spaces because of how this country treats us. 💔
17
57
282
16,042
Rest in peace @6e726d This was the last time we got to hang out. From the early days of my WiFi research he quietly supported and advised me - super smart and super humble. You’ll be missed man.
2
5
70
6,576
quantumhello.xyz/?q=https%3A…
My silly little C# WebServer now has post-quantum encryption and partial h2 support.
On a side note - h2 is weird. It's all single-connection binary streams instead of plain text.
Also need to update the code on the Github some day...
1
1
3
276
I added post-quantum authentication (ML-DSA) checks to QuantumHello, thanks to the ML-DSA support in go 1.27 from @FiloSottile and others. quantumhello.xyz
6
869
Yep.
We're publishing our most detailed threat intelligence report to date.
It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them.
We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies.
These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve.
We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop.
Read the report: anthropic.com/threat-intelli…
2
4
3,391
