A Primer on Subverting AI/ML
Before placing implicit trust in AI/ML models for decision-making it is worth considering how they can be compromised. Here are some ideas.
Writing about nuclear security, technology, systems thinking, and the cool things I build.
Mitchell Hewes Read more about why
Before placing implicit trust in AI/ML models for decision-making it is worth considering how they can be compromised. Here are some ideas.
Protecting against malicious acts is complex. Does the focus need to move from threats to function-based protection against consequences?
Fun fact: my first work experience was with maps.
Wherein I post pictures of warmer climates.
Worth it.
C'était la première fois que je voyais la tour Eifel ; elle ne m'a plus jamais regardé de la même façon.