YouTube announced this week the rollout of “Thumbnail Test & Compare,” a new tool for creators to see which thumbnail performs the best.
When creators publish a new video (or update an existing video), they can test up to three thumbnails, which will be shown “evenly” across viewers, the company explained in a post.
YouTube encourages testing thumbnails with distinct differences, such as variations in layout compositions, backgrounds, and text overlays, to ensure a comprehensive test.
Creators can test thumbnails on a variety of content, including regular videos, podcast episodes, archived livestreams, and public long-form content.
Over the coming weeks, Thumbnail Test & Compare will become available in YouTube Studio on desktop to all creators with access to advanced features.
But to reuse, you have to recover, and SpaceX is proving that it will be able to do just that with Starship.
The ultimate goal is to fly Super Heavy and the Starship upper stage back to Starbase, SpaceX’s private Starship launch and development site in southeast Texas, where they’d make vertical landings on solid ground.
Starship lifted off from Starbase at 8:50 AM CT, the fourth launch in the rocket test campaign.
A little over an hour after launch, Starship followed suit, surviving the extreme heat from traveling through the Earth’s atmosphere at hypersonic speeds and splashing down in the Indian Ocean.
During that test, SpaceX also tested capabilities that will be key for delivering payload to space, including opening and closing the payload door.
The self-driving technology company announced Wednesday plans to begin testing in Austin and Miami this summer.
Earlier this week, the National Highway Traffic Safety Administration requested more information from Zoox to aid its probe into rear-end crash risks posed by unexpected braking.
Zoox also didn’t say when it aims to remove the safety driver or begin commercial operations in Austin or Miami.
The plans to test in Austin and Miami come as Zoox gears up for its first commercial launch.
The company won’t be testing those vehicles on public roads in Austin or Miami, just yet.
Last week, Meta started testing its AI chatbot in India across WhatsApp, Instagram, and Messenger.
Meta confirmed that it is restricting certain election-related keywords for AI in the test phase.
When you ask Meta AI about specific politicians, candidates, officeholders, and certain other terms, it will redirect you to the Election Commission’s website.
But just like other AI-powered systems, Meta AI has some inconsistencies.
This week, the company rolled out a new Llama-3-powered Meta AI chatbot in more than a dozen countries, including the U.S., but India was missing from the list.
As I wrote recently, generative AI models are increasingly being brought to healthcare settings — in some cases prematurely, perhaps.
Hugging Face, the AI startup, proposes a solution in a newly released benchmark test called Open Medical-LLM.
Hugging Face is positioning the benchmark as a “robust assessment” of healthcare-bound generative AI models.
It’s telling that, of the 139 AI-related medical devices the U.S. Food and Drug Administration has approved to date, none use generative AI.
But Open Medical-LLM — and no other benchmark for that matter — is a substitute for carefully thought-out real-world testing.
Shares are not continuing to climb in early trading, but are holding steady above its IPO price, at around $100 at the time of writing.
Its successful debut marks the third major tech IPO in the United States this year, and is the third in a row to price well and immediately trade higher.
Investor eagerness for Ibotta indicates that “there is an increasing appetite for IPOs again” Smith said, “particularly in the tech space.”Don’t pop the champagne yet for the tech IPO market coming roaring back, however.
Classic tech IPOs tend to feature tech companies still in growth mode and deeply in the red.
Smith agrees, calling the upcoming Rubrik IPO “an even bigger test” for tech debuts “given its weaker current financial picture.”We’ll find out next week.
Inversion Space is aptly named.
Inversion has developed a pathfinder vehicle, called Ray, that’s a technical precursor to a larger platform that will debut in 2026.
Impressively, the company has designed and built almost all of the Ray vehicle in-house, from the propulsion system to the structure to the parachutes.
“The purpose of our Ray vehicle is to develop technology for our next-gen vehicle.
As such, we’ve built basically the entire vehicle in-house,” Fiaschetti said.
Meta-owned social network Threads is finally testing a “Recent” filter to sort search results by the latest.
“We’re starting to test this with a small number of people, so it’s easier to find relevant search results in real-time,” he said in a reply to a user.
A user part of the test posted that they could see “Top” and “Recent” filters on the search results screen.
They noted that the “Recent” filter isn’t strictly chronological, but it shows the latest posts better than the “Top” filter.
Earlier this year, the company accidentally rolled out the option to sort search results by the latest.
Smart ring manufacturer Oura is introducing a new section in its app called Oura Labs to test out new features and get user feedback.
Oura said that Symptom Radar will monitor biometric signals such as body temperature range, respiratory rate, resting heart rate, and heart rate variability.
“Oura Labs is our approach to recreate internal engagement for new features with users in a structured and formal way.
Users will get to see a lot of early-stage ideas in Oura Labs,” Patel said.
Users can provide direct feedback about these experimental features along with general feedback for the product through Oura Labs.
Founder Daudi Barnes started the company in 2019 to augment the work of his previous company, Advanced Mobile Propulsion Test.
The Colorado-based startup already operates one test stand, called Sunshine, which AMPT stood up in 2010.
“The market is just really, really expanding really fast right now,” Animas project manager Graham Dudley explained.
So they’ve had four plus years of design and development that they have to reboot on, and that’s really, really expensive and hard for your schedule prediction.
Its a competitive edge in the space propulsion market, which has become increasingly crowded as the cost to launch spacecraft to orbit has dropped.