Florian DietzinTowards Data ScienceAchieving AI alignment through deliberate uncertainty in multiagent systemsMaking an an AI safer for humans, via deliberate deception, manipulation and confusion.Jun 11, 2020Jun 11, 2020
Florian DietzinThe StartupMind hacks based on machine learningWhere AI research overlaps with psychology, you can learn interesting mind hacks and productivity tricks.Apr 10, 2020Apr 10, 2020
Florian DietzinTowards Data ScienceSelf-assessing neural networksGiving neural networks the ability to assess the reliability of their own features, and to prioritize features based on context.Nov 21, 20191Nov 21, 20191
Florian DietzinTowards Data ScienceWhy your AI might be racist and what to do about itRacism among humans is a social problem, but racism in an AI is an engineering problem.Nov 9, 20192Nov 9, 20192
Florian DietzBreaking the Hardest Logic Puzzle Ever for fun and profitHow to go from “This is a really hard logic puzzle” to “Let’s enslave a god and achieve omnipotence by asking a really tricky question”.Aug 2, 2019Aug 2, 2019
Florian DietzinThe StartupA new crowdfunding mechanism for tech startupsI present a new method for crowdfunding tech startups, with a unique incentive structure to capitalize on exponential growth curves.Jul 30, 2019Jul 30, 2019
Florian DietzinTowards Data ScienceAutomation in Data ScienceWhy the most technical parts of my work keep getting easier, and the most difficult parts have nothing at all to do with AI.Jul 20, 20192Jul 20, 20192