Nick Hutton (@nickdothutton) header
Nick Hutton (@nickdothutton) profile picture

30 years in Cyber Security, Product, Investor, Founder, Engineer. Armchair interests: geopolitics, propaganda, persuasion, ideological subversion.

The Cyber-Industrial Complex
Joined November 2016
I wrote about this problem in 2019 and before, as part of my "Winning Systems" series of posts. Attack Surface Reduction, and "How the cave fish list his eyes".
1
1
9
Britain’s primitive tribes, government, hospitals, schools, businesses, stagger about. Confused, fearful, struck dumb beneath a dark sky. As if the sun failed to rise, swallowed by forces they could not name. Email and calendar gone, silenced by some cosmic power beyond their comprehension. UK Resilience.
55
Replying to @SO3_Clausewitz
Late 1950s to early 60s was the best period of the last century. 30% real wage growth, real disposable income rose by over 22%, moderate inflation. Culture etc.
3
736
We don't even need AI to tell us how bad software is. Static analysis reveals it, profiling reveals it. Most software is easy meat for AI. Assume vulnerable. Act accordingly. This is the work.
I won’t pretend to understand the ultimate goal of investing so heavily into making models good at vulnerability research (and not any number of defensive competencies in cyber). The way model proficiency in VR is being used is primarily to substantiate a view that the models are dangerous, problematic, and in need of regulation. I could come up with a plausible narrative that has the labs recognizing the ephemeral value of frontier model training (heavy investment for short term returns) and emphasizing the products dangers as a bid towards regulation as a strategic necessity to limit competition. An ok marketer could argue that fearing the product implies how formidable it is (in the short-term). But that all feels too contrived. A listener (former exec at a big company I’d often comment on) recently pointed out that despite my best efforts, I still fall prey to the delusion that organizations engage in strategic decision-making when it’s mostly incompetence, intransigence, and whim. In response to that apt observation, I’d give more weight to following: The labs have focused on producing models disproportionately capable at vulnerability research due to a combination of it being (1) easy, (2) high-profile, (3) commercially superfluous, and (4) central to the doomer narrative. 1. Vulnerability research appears easier to RL than the rest of cybersecurity. 2. VR is a target-rich environment for milestones to publicize. Even if the models hadn’t gotten this good, there’s a wealth of vulns to hype for messaging. 3. VR competency doesn’t yield an immediate product with a quantifiable value (don’t confuse this with the exploit business). So, no comparables or preset price bands. No need for enterprise adoption or organizational diffusion before claiming success. 4. The competency lends itself to hyperbolic and imaginative myth making about cyber ‘weaponry’ that seemingly validates doomer narratives and tropes (alignment, containment) while gaining fast traction with policy folks and natsec types swayed by militaristic analogies. I wouldn’t call it an effective strategy because it isn’t one. Hearing that the labs lack the internal coherence to productize anything effectively, byproduct of circumstance seems the more likely origin story.
1
53
"Well minister, you see the precise wording of questions ensures you don't get the wrong answer."
1
33
As usual, former GCHQ Directors tend to speak more plainly than those currently in post. 🎙️from @mccloskeybooks & @gordoncorera. youtube.com/watch?v=kE_jmF2D…
1
67
When the UK was a high trust society, there was something called the "Luggage Ahead" system (or Passengers' Luggage in Advance), alongside the normal use of open luggage vans. The whole thing ran on trust.
The biggest mistake tourists make in Japan is dragging a fifty-pound suitcase up and down subway stairs. Watch how Japanese travelers board the bullet train. They walk on with a cup of green tea, a small shoulder bag, and both hands free. Where is their luggage? It is already in the next city, waiting for them. In Japan, the system is called Takkyubin, recognizable everywhere by a logo of a mother black cat carrying her kitten gently in her mouth. When you check out of your hotel in Tokyo at eight in the morning, you hand your heavy suitcase to the front desk, fill out a tiny slip, and pay about twenty dollars. You can even send it straight from the corner convenience store. Then you spend the entire day wandering through mountain temples or eating street food with nothing on your back. When you walk into your traditional inn in Kyoto that evening, your suitcase is already standing neatly in your room, wiped clean, with the handle turned toward you. In fifty years of handling millions of bags a year, lost luggage is virtually unheard of. A society runs smoothly when you can hand your belongings to a stranger at a convenience store and know they will beat you to your hotel.
2
1
7
741
Replying to @ecurrnomics
Log scale [check] Massaged Categories [check] Standardisation that obscures specificity [check]
3
2
35
2,479
Replying to @AndrewOrlowski
The only good thing about that development is that it will begin to collapse shortly after the NHBC expires.
64