Since some people seem to not be familiar with Joanna, she is a massively influential security researcher. IMO most famous for her "Blue pill" attacks on Vista and Xen: https://blackhat.com/presentations/bh-usa-06/BH-US-06-Rutkow..., https://invisiblethingslab.com/resources/bh08/part3.pdf. Her work demonstrated that hardware virtualization is not in fact the security panacea we wish it was, but that it too is vulnerable to attack just like any other layer of the stack. This revelation resulted in significant changes in the security world.

She is also famous for the Qubes OS project and coining the term "evil maid attack".

Is there some important context I am missing?

In case you read this: Hi Joanna! Nice to hear from you. Looking forward to your thoughts and sharing.

Why did the author leave computer security industry?

> This blog is intended to be a reportage of my struggles between Rationality and Humanism. Pragmatism and Beauty. Formalism and Intuition. Freedom and Love. Individualism, Independence and Privacy on the one side, vs Egalitarianism, Sharing and Community on the other.

What is he on about? Sounds like his future blog posts could just be rambling about anything.

Well known security researcher resurfaces, first blog post mentions Claude:

https://tracesofhumanity.org/freediving-embodiment-and-human...

Please, please, please let this not be the incentive!

Joanna, if your reading this, welcome back .. really miss your blog !

In my ideal future, Joanna would write about things like :

  - how do we secure LLMs leading up to AGI, presumably they can bypass our best attempt at airgap firewalls
  - is consciousness different from thinking from intelligence ?
  - how do we ensure AI is distributed so all humans can benefit, and not used as a multiplier of extreme wealth inequality ?
  - whats missing in current AI LLMs ?

I'll take a crack at the last one : imo, the current generation of LLMs is missing :

  - a bias for truth / fact .. or a mode switch to make it bias truthiness
  - reasoning by chains of formal language [ formal logic ]
  - reasoning by chains of probabilistic inference [ bayesian logic ]
  - reasoning by deep simulation [ stochastic modelling ]
  - spatial reasoning [ 3D model of space, machinery, physics. 3D reconstruction. model of humans and animals with bones, muscles, mass ] 
  - mathematical modelling [ proposing formulae, checking fit ]
  - psychological model of humans, of human populations, their needs / motivations / rewards and psychoses

Essentially AGI would require a proper merging of RL style [ NN learning from a stochastic simulation of future states ] and current ChatGPT style LLMs.

An implication of the above is that future AGI will need to run on fast branching CPU _and_ massively parallel GPU with a fast data path between them - ie. balanced compute.

However, the best part will be discovering what the real Joanna writes about !