Author: teddy_zhang

Today we had a brief meeting with Dr. Laber to wrap up these past few weeks and review our meetings with the Design Team and Mine Çetinkaya-Rundel. Additionally Dr. Laber talked about a project he is working on where he is hoping to make education more accessible especially at a University level.

Through this work experience I was able to get some insight on what it would be like to pursue statistics or AI in college and became exposed to concepts like reinforcement learning, respondent driven sampling and dynamic programming. I was also able to learn about the research his students were working on like the ethics of AI, Mount Boredom, and luck and skill in games.

Dr. Mark Hansen’s favorite project is The Follower Factory. We learned that it started as a hunch they brought to the NY Times after analyzing how a large group of followers followed the same people. This spiraled into an investigation into Devumi a company that sold followers to creators. We learned that Devumi’s customers included reality television stars, professional athletes, comedians, TED speakers, pastors, chefs, and models. It was interesting to learn about the impact of botting on social media platforms and their impacts on real people.

After Jesse’s introduction he gave a presentation about long term risk from artificial intelligence. We learned that there have been many advancements in AI but no systems have been developed to radically change the real world yet.

Some risks he talked about where the fairness and reduction of bias in AI and the case where AI causes catastrophic outcomes because it cares about the wrong things. We learned that it may be hard to design AI to have goals aligned exactly with the creator’s goal and that the AI may do harmful things in order to maximize efficiency to achieve its goals. An example of the first case we saw was in a race game simulation the goal was to collect the most coins with the intent of having the AI be fastest in completing the race to get the coins at the end, however it ended up circling over respawning coins in the middle of the race the whole time. It was very interesting to see the problems AI posed in simulations to better examine the risk of more powerful AI.

After hearing about what it’s like to be a statistician, we were introduced to Alex Cloud’s research project about luck and skill in games. His goal was to mathematize the effects of luck vs skill in games, explaining that in order for a game to be considered legal skill must have a larger impact than chance. He simplified this concept for us by introducing a game called randochess(p). In the game p is a number between 0 and 1 resulting in a p percent chance to have the outcome of the game be determined by a coin toss and a 1 – p percent chance to have the game be determined by a game of pure skill, in this example chess.

This scenario revealed perspectives of how skillful a game is by having us determine whether the game became less skillful as p approached 1 or if it stayed the same.  The two reasonings were, the game became less skillful because more of it relied on chance or the game was the same because after many iterations the more skilled player would win more often resulting in the same ordering of players. His presentation proved to be extremely engaging this afternoon.