Before batteries and household electricity were used to power clocks, most clocks had to be wound by hand to keep operating. Eight-day clocks were designed so they only had to be wound every eighth day and the movement only turned in a clockwise direction. Therefore, someone with an appearance objectionable enough to stop the clock and send the movement spinning in the wrong and opposite direction would be ugly indeed.
All of which is why the conclusions of this manifesto are precisely backwards. It’s true that women are socialized to be better at paying attention to people’s emotional needs and so on — this is something that makes them better engineers, not worse ones. It’s a skillset that I did not start out with, and have had to learn through years upon years of grueling work. (And I should add that I’m very much an introvert; if you had asked me twenty years ago if I were suited to dealing with complex interpersonal issues day-to-day, I would have looked at you like you were mad.) But I learned it because it’s the heart of the job, and because it turns out that this is where the most extraordinary challenges and worthwhile results happen.
Yep. Nothing impresses a 9 year old as much as drawing their favorite cartoons.
Most annotation tools avoid making any suggestions to the user, to avoid biasing the annotations. Prodigy takes the opposite approach: ask the user as little as possible, and try to guess the rest. Prodigy puts the model in the loop, so that it can actively participate in the training process and learns as you go. The model uses what it already knows to figure out what to ask you next. As you answer the questions, the model is updated, influencing which examples it asks you about next. In order to take full advantage of this strategy, Prodigy is provided as a Python library and command line utility, with a flexible web application. There’s a thin, and optional hosted component to make it easy to share annotation queues, but the tool itself is entirely under your control.
Sure. Teach the machines to teach themselves. Next thing you know they’ll take low paying thankless jobs away from our public school teachers.
This raises an interesting legal question: Is it a crime to create and sell malware?
So wait, does that make it a crime to create and sell something for the purpose of harming another? I can think of a large American industry that should be worried about that precedent.
Though Shkreli is best known for hiking the price of a decades-old drug by 5,000 percent when he led Turing Pharmaceuticals, this trial related to an entirely different matter: Ponzi-like fraud. Specifically, Shkreli was accused of eight counts of securities wire fraud linked to his behavior at two hedge funds and at a previous biopharma company he led, Retrophin Pharmaceuticals.