For anyone spending over an hour a day in their inbox, tools that improve the experience are a huge deal. The fancy new email product offerings vary widely, because each was founded upon unique ideas of what email can and should be. The following are the biggest differences I identified after using HEY for 8 days, Superhuman for 9 months, and Gmail for 15 years.
Genitalia was scribbled on my shared screen within the first minute of an event about building products. At one point, a participant began twerking, another undressed, and a few others pasted hundreds of hateful lines in the chat. One Zoom bomber loudly voiced a fake apology. Some of these disruptions were delayed tens of minutes into the event by sleeper cells of sorts. During each new form of disruption, we scrambled to disable another Zoom feature, and remove more attendees.
Our group of event organizers were embarrassed, shaken and perplexed. The trolls won this round. In contrast, I visited a professional Zoom with 200 strangers a week earlier in which all attendees could unmute freely, yet there was no foul play. So I was confused as to how my meeting got so bad so fast. …
Kaggle is home to the best and worst of machine learning. It’s the battle grounds for thousands of teams vying to build the best models, solve real-world problems on a deadline, with hundreds of thousands of dollars at stake. Training these world-class models on large datasets often amounts to hundreds of compute hours. In this post, we’ll explore some techniques winning teams use and how to speed things up.
Sometimes competitions on Kaggle result in models that are impossible to reproduce or are out-right fraudulent. But some of these competitions set new bars for the entire industry. Kaggle is where machine learning is in the spotlight and huge strides of progress are made. …
Kaggle just announced that the 1st Place Team, Bestpetting, has been disqualified from the Petfinder.my competition for cheating. The team crawled the pet adoption site to collect the private leaderboard answers, and hid this data in their submission to win the first prize on 2019–04–09. The first prize was $10,000 out of the $25,000 prize pool. In this post, you’ll find background on Kaggle, the darker side of its competitions, and links to the related resources.
Kaggle (a subsidiary of Google), is an online community built around competitions to build machine learning models. With prize pools as high as $1,500,000, the platform attracted a diverse following. Such competitions present a dataset, and the metric which will be used to decide the winning submission. Competitors analyze the given data, build models to match the desired outcome, and submit their results (often alongside their code). To prevent cheating, machine learning competitions include data that isn’t labeled and is used in two…
In my career there weren’t many times when I found a tool that felt like a leap in productivity. I remember first discovering Python as a C software engineer, browsing an open source project on GitHub, and collaborating on a movie script with Google Docs. Notion is as big of an upgrade in the following capacities.
To quote Wikipedia:
A race condition or race hazard is the behavior of an electronics, software, or other system where the system’s substantive behavior is dependent on the sequence or timing of other uncontrollable events.
// When you run this script using Node or in a browser, it
// does not print "Ended with 0", but a random number.
// Even though the functions running
// simply loop 100 iterations of adding and subtracting.
// The reason the end result is random is because the
// sleeps are of random duration and the time between the read
// of the variable causes the eventual write to be incorrect
// when `adder` and `subber` interleave.
// This problem is similar to:
On 2019–07–13 a group of talented VFX artists named Corridor Crew posted a DeepFake video of Keanu Reeves without any labeling that it’s fake. It seems Keanu’s identity and brand is being taken advantage of without his consent, and the audience is intentionally being mislead. The only minor mention of the ruse is in the description “This video was created using Deepfake technology”. That does not at all clarify that Keanu Reeves took no part in this production.
Previously, the Corridor Crew created a video titled “We Made The Best Deepfake on The Internet”, clearly explaining it only featured an impersonator of Tom Cruise. This new video titled “Keanu Reeves Stops A ROBBERY!” is clearly staged, but almost every piece of content that Keanu is involved in is staged. …
We just shipped a feature called “Shared Data Storage” and it’s pretty useful if you find yourself attaching different metadata to the same underlying data, or saving multiple datasets with a lot of overlapping content.
For example, imagine storing 100,000 chest x-ray images weighing in at 43GB zipped. That number isn’t a big deal today considering a 1 TB drive costs about $100. But storing this amount of data efficiently can be challenging when you consider a whole team needs access to it, the data changes over time, and that you’d like to keep track of these changes. In a computer vision company — datasets are always evolving with data streaming from production, labeling and analysis. …