Inside the man against the hackathon machine

Then there Eric Chung37 -year -old and has a background in dentistry and previously copied an emerging company that simplifies the medical bills of dentists. It was placed on the “machine” team.

“I will be honest and say I am very comfortable to be on the machine team,” says Chung.

In Hackathon, Chung was building programs using sound and face recognition to detect autism. Of course, my first question was: Not there is wealth Who is the problems with this, like a biased data that leads to false positives?

“A short answer, yes,” says Chung. “I think there are some wrong positives that may appear, but I think it is with sound and in the face of the face, I think we can improve the accuracy of early detection.”

Agi ‘Tacover’

The workspace, like many AI’s things in San Francisco, have relationships with effective altruism.

If you are not aware of the movement through Bomb fraud addressesIt seeks to increase the good that can be done using the time of participants, money and resources. The next day of this event, the event space hosted a discussion on how to take advantage of YouTube “to connect important ideas such as why people eat less meat.”

On the fourth floor of the building, the bulletins covered the walls-“AI 2027: Will Agi Tacover” showing a recently passed taku bulletin, not another providing “pro-animal exercise”, that is, another context.

Half an hour before the applicant, programmers cheated submarines of IKE’s vegetable meat and rushed to finish their projects. One floor below, judges began to reach: Brian Voca and Xiamal Hitch Annakat From Openai AI applied, artificial intelligence, Marius Bolandra From the Applied Applied Intelligence Team, and Varine NairEngineer from the start of Amnesty International factory (It is also the process of the event).

With the referee, a member of the Metr, Nate Rush team, the Excel table that tracked the contestants, showed me with groups of green and human colored projects in red. Each group moved up and down the list when the rulers entered their decisions. “Do you see it?” He asked me. No, I don’t – Mishmash did not appear in the colors any clear winner until half an hour in the ruling. That was his point of view. To surprise everyone, the man for the machine was closely racing.

show time

In the end, the final candidates were divided evenly: three from the “man” side and three “machine”. After each experimental offer, the crowd was asked to raise their hands and guess whether the team used artificial intelligence.

First Up was Wiewsense, a tool designed to help people with visually impairment in their surroundings by copying direct video assistants to a text of the screen reader to read loudly. Looking at the time of short construction, it was technically impressive, and 60 percent of the room (through EMCEE) believed that it uses artificial intelligence. He did not.

After that, the team that built a platform for web design was using the pen and paper, using a camera to track actual time – did not participate in the coding process. The pianist project is presented to the finals through a system that allows users to download piano sessions for comments created from artificial intelligence; It was on the side of the machine. Another team displayed a heat maps tool for code changes: critical safety problems appear in red, while routine modifications appear in green. This one used artificial intelligence.