AI in Cyber – Challenges and Solutions
Jamie Graves from ZoneFox gives an insight into how we can use AI more effectively. It was standing room only when Jamie Graves of ZoneFox took to the stage at AI-Analytics X to discuss the strengths and weaknesses of AI.
Jamie started with some of the more nonsensical examples of AI. From computers identifying sheep as flowers to viewing clouds as sheep. These examples demonstrated that AI has some unintended consequences and, in some cases, still requires work.
Sharing more life-impacting examples, Jamie revealed cases where offenders have been sentenced based on algorithms generated by AI. Based on either known or unknown capabilities, it is a daunting prospect that the fate of offenders hangs in the balance of a computer not having the full insight into a situation.
Rubbish in equals rubbish out Known as the Black Box Problem, decisions determined by the data put in generate the result that is pushed out. The age-old phrase ‘garbage in = garbage out’ rings true, where it’s necessary to craft data in the appropriate way to ensure the correct algorithms are generated.
Jamie went on to break up the types of AI and described the challenges and benefits these provide:
How do we know the AI in our systems hasn’t been tricked to falsely come up with a solution? The closest human comparison is a visual allusion, which tricks our minds, and it’s important for computers to see data for what it is, rather than making assumptions.
This highlights the requirement to ensure data collected and captured is suitable for the needs of the cause and to ensure it is useful. This is where the role of a data scientist is becoming more pertinent to ensure the data is being captured and analysed in the most appropriate way.
Data scientists are becoming more prominent in organisations to accurately identify the story the data is telling for this exact purpose.
Discover what Jamie’s session covered in regards to the human factor, and learning quickly and smartly by downloading the full eBook here.