My recent trip to Alaska was an eye-opening experience on many levels. The natural beauty of the state was breathtaking, and I was fortunate to meet some amazing people both on and off the ship. However, I was most struck by the challenges some senior people faced, who traveled in their walkers, wheelchairs, and scooters.
One of the most frustrating things I saw was how the cruise’s slow and buggy app made ordering food difficult for these passengers. They struggled to navigate the app and often had to wait for a crew member to help them. This was not only inconvenient, but it was also physically taxing for these passengers.Despite these challenges, I was impressed by the willingness of the cruise staff to accommodate the needs of their passengers with challenges. They were always patient and helpful, and they went out of their way to ensure everyone had an enjoyable trip. My experience in Alaska taught me a valuable lesson about the importance of accessibility. I will keep this experience in mind going forward, and I will do my part to make the world a more accessible place.
The article says it will be a small release, but I am still happy about the new Change Array by copy features: toReversed(), toSorted(compareFn), etc.
Cost-effective tracing with sampling
I’ve sampled my metrics and traces but was unaware of these terms. Head-based sampling makes a sampling decision before a request is handled, usually based on a sampling ratio. Tail-based sampling makes a sampling decision based on how the request was handled (status code, response time, etc.)
dalai: The simplest way to run LLaMA on your local machine
After two commands and an hour-long wait, I was running a language model with an accompanying web app on my laptop.
People ❤️DevEx: What Actually Drives Productivity
This study supplements DORA metrics by adding and categorizing 18 metrics to monitor developer experience. Three categories are feedback loops, cognitive load, and flow state, each with six metrics. My org can benefit from measuring these metrics to align our efforts.
Whose Opinions Do Language Models Reflect?
I’ve read about biases in machine learning models, but they’ve been anecdotal. But this study measured the bias by comparing the responses from language models to the public poll results. It found that the models are more left-leaning than the general public in the US.
Business 💰100 things we announced at I/O 2023
Of the hundred things, here are the five that stood out to me:
1. We introduced the Google Pixel Fold to our Pixel portfolio!20. The [generative AI search] experience will take your search queries and give you AI-powered snapshots of key information to consider with links to dig deeper and suggested next steps26. Over the last year, AI features in Workspace helped users write over 180 billion times.52. We announced a handful of improvements to and news about Bard, our experiment that lets you collaborate with generative AI76. Previously only available to Google One subscribers in the U.S., we’re expanding access to our dark web report over the next few weeksInteresting Finds 💡The case of the 500-mile email
A fun story about how light can only go 500 miles in 3 milliseconds.