Trailer: Learn how to hack neural networks, so that we don't get stuck in the matrix!

Summary

In this talk, Johann introduces an exciting opportunity to learn how to build and break machine learning systems. He emphasizes a hands-on approach to understanding neural networks, exploring how to threat model and hack machine learning systems. The session promises practical examples and an engaging learning experience. Johann invites attendees to join him at Gray Hat for a fun and educational exploration of AI and hacking, encouraging everyone to dive deep into the world of neural networks and their vulnerabilities.

Takeaways

• 😀 The speaker invites people to learn how to hack a neural network and avoid getting stuck in the 'Matrix'.
• 😀 The focus of the talk is on building and threat modeling machine learning systems.
• 😀 Attendees will gain practical, hands-on experience during the talk.
• 😀 The talk is designed to be fun and educational, with plenty of real-world examples.
• 😀 The speaker's name is Johann, and he expresses excitement about meeting the audience.
• 😀 The event is hosted at 'Gray Hat', a venue where security and hacking discussions are central.
• 😀 The title of the talk is 'Learning by Doing: Building and Breaking a Machine Learning System'.
• 😀 The talk emphasizes the importance of understanding both building and attacking machine learning systems for security purposes.
• 😀 The speaker promotes a practical, immersive approach to learning about machine learning and security.
• 😀 The speaker ends with a friendly and welcoming tone, expressing hope to see many attendees at the talk.

Q & A

What is the main topic of Johann's talk?

Johann's talk is about learning how to build and break a machine learning system, with a focus on practical examples and threat modeling.

What is the purpose of the talk at Gray Hat?

The purpose of the talk is to teach attendees how to approach machine learning systems from both the building and breaking perspectives, offering hands-on learning experiences.

What does Johann mean by 'learning by doing'?

'Learning by doing' refers to gaining knowledge through practical, hands-on experience rather than just theoretical understanding, specifically in building and breaking machine learning systems.

What is meant by 'threat modeling' in the context of machine learning?

Threat modeling in machine learning involves identifying potential vulnerabilities and risks within the system to anticipate and mitigate attacks or issues before they arise.

Why is it important to understand how to break a machine learning system?

Understanding how to break a machine learning system is crucial because it helps in identifying weaknesses and vulnerabilities, which can then be fixed to improve security and reliability.

What type of audience is Johann's talk intended for?

Johann's talk is intended for individuals interested in machine learning, ethical hacking, and security, particularly those who want to gain practical experience in these areas.

What can attendees expect from Johann's talk?

Attendees can expect practical examples of building and breaking machine learning systems, along with a fun and engaging learning experience.

How does Johann describe the learning experience at Gray Hat?

Johann describes the learning experience as fun and interactive, emphasizing practical examples that make the concepts easier to grasp.

What is the significance of 'Gray Hat' in the context of Johann's talk?

Gray Hat refers to the conference or event where Johann will be giving his talk, which is focused on hacking, security, and machine learning systems.

What joke does Genie, the funny hacker, tell?

Genie jokes: 'You ever heard of the neural network that walked into a bar? Yeah, it got caught in a feedback loop and kept asking for a byte.'