Kalakshar US

OpenAI is rolling out a new partnership program to collect datasets from third parties that it intends to use to train its AI models. The initiative, OpenAI Data Partnerships, will seek large-scale private and public information that it says is “not already easily accessible online to the public.” The company says the data it will collect doesn’t necessarily have to be quantitative or in text formats — the program will also accept images, audio or video.

Notably, the company says it’s on the lookout for data on “any topic” and in “any language” so long as it “expresses human intention,” which it likens to long-form essays or transcribed conversations. Human-centric data collected by OpenAI is expected to help the company improve tools like its automatic speech recognition technology which is used to transcribe spoken words. This initiative also lines up with ChatGPT’s recent expansion to support voice queries to engage with users in a conversational manner. Exposing its AI models to more information that teaches it how to hold up human-like conversations will only further improve this feature and other tools that will follow in function.

The model testing conducted throughout the data partnership program will also naturally expand the capabilities of OpenAI’s consumer-facing GPT-4 Turbo, which has been updated to provide users with more complex and meaningful responses. OpenAI says it has already started working with interested organizations, including authoritative bodies like the Icelandic government. Through curated datasets, OpenAI says its working to improve GPT-4’s ability to comprehend queries made in the Icelandic language.

If a private or public organization wants to participate in the program, a representative can submit a form on the company’s website and share information on the data type and size that they intend to share. There are two pathways for datasets. The first is the Open-Source archive, which is ideal for datasets relevant to training language models. However, submissions made to it will be public for anyone to use. Alternatively, OpenAI says a company can submit information through its private dataset pathway which will be funneled to train proprietary AI models, which the company says includes their “foundation models” and “fine-tuned and custom models.” This is recommended for companies or institutions that want to keep their data confidential. But in that same regard, OpenAI says it is not looking for datasets that contain sensitive or personal information.

ChatGPT has already set records for its soaring user base. It has about 100 million weekly active users around the world, meaning privacy will only continue to be a focal point for the tool. Previously, Samsung employees were put in the hot seat for leaking sensitive data to the AI model. While OpenAI claims it does not use data generated by its API to train its models unless a user explicitly submits information through an opt-in form, all eyes will be on how the company handles the data collected through this initiative, especially the private datasets.

This article originally appeared on Engadget at https://www.engadget.com/openai-wants-to-work-with-organizations-to-build-new-ai-training-datasets-214548902.html?src=rss 

Tumblr, a flailing social media site from a bygone era, may be run by a skeleton crew from now on. An alleged internal memo from parent company Automattic has made the rounds on social platforms (including Threads), stating it has “not gotten the expected results from our effort.” The decision appears to mark a sharp U-turn from a separate leak this summer, claiming Automattic was building a TikTok-like algorithmic feed into the aging site.

Although this doesn’t quite appear to be the end of the road for Tumblr, the note doesn’t sound promising for the platform’s future. It says “the majority of the 139 people” will switch to other Automattic projects, leaving a barebones gang of Trust & Safety and support workers to oversee Tumblr’s smoldering embers. Given how many brutal layoffs we’ve seen this year, handling the transition in a way that avoids job losses could be a silver lining.

Automattic, the company behind the blogging tool WordPress, acquired Tumblr in 2019 from Verizon, which landed the platform through its purchase of Yahoo! (Engadget’s parent company) in 2017. It likely didn’t help that its ownership turned into a game of musical chairs, and none of them seemed to find the right formula to get the microblogging network back on its feet. (Its controversial ban on adult content likely had something to do with that.)

“We are at the point where after 600+ person-years of effort put into Tumblr since the acquisition in 2019, we have not gotten the expected results from our effort, which was to have its revenue and usage above its previous peaks,” the alleged memo reads. After throwing in cliches about climbing mountains and being better to try and fail than not to try at all, the note claims the team’s next step is to “reflect and decide where else we should concentrate our energy together.”

Engadget reached out to Automattic for comment and confirmation but didn’t immediately receive a response. We’ll update this article if we hear back.

In addition to WordPress, Automattic’s other brands include the journaling app Day One, the e-commerce plugin WooCommerce, Gravatar and the note-taking app SimpleNote.

This article originally appeared on Engadget at https://www.engadget.com/tumblrs-staff-is-reportedly-reduced-to-a-skeleton-crew-215853169.html?src=rss 

One of the major benefits of certain artificial intelligence models is that they can speed up menial or time-consuming tasks —- and not just to whip up terrible “art” based on a brief text input. University of Leeds researchers have unveiled a neural network that they claim can map an outline of a large iceberg in just 0.01 seconds.

Scientists are able to track the locations of large icebergs manually. After all, one that was included in this study was the size of Singapore when it broke off from Antarctica a decade ago. But it’s not feasible to manually track changes in icebergs’ area and thickness — or how much water and nutrients they’re releasing into seas.

“Giant icebergs are important components of the Antarctic environment,” Anne Braakmann-Folgmann, lead author of a paper on the neural network, told the European Space Agency. “They impact ocean physics, chemistry, biology and, of course, maritime operations. Therefore, it is crucial to locate icebergs and monitor their extent, to quantify how much meltwater they release into the ocean.”

Until now, manual mapping has proven to be more accurate than automated approaches, but it can take a human analyst several minutes to outline a single iceberg. That can rapidly become a time- and labor-intensive process when multiple icebergs are concerned.

The researchers trained an algorithm called U-net using imagery captured by the ESA’s Copernicus Sentinel-1 Earth-monitoring satellites. The algorithm was tested on seven icebergs. The smallest had an area roughly the same as Bern, Switzerland and the largest had approximately the same area as Hong Kong.

With 99 percent accuracy, the new model is said to surpass previous attempts at automation, which often struggled to tell the difference between icebergs and sea ice and other features. It’s also 10,000 times faster than humans at mapping icebergs.

“Being able to map iceberg extent automatically with enhanced speed and accuracy will enable us to observe changes in iceberg area for several giant icebergs more easily and paves the way for an operational application,” Dr. Braakmann-Folgmann said.

This article originally appeared on Engadget at https://www.engadget.com/a-neural-network-can-map-large-icebergs-10000-times-faster-than-humans-212855550.html?src=rss 

The Spotify TV app has been revamped to make the big-screen app experience resemble the popular mobile version of the digital music and podcast streaming service. The home page will have a “Now Playing View” and feature shortcuts to recently played content. The TV app will also recommend new songs and podcasts based on your listening history on the main screen, similar to what you might get on the mobile app.

The TV app home page doesn’t just look different. You can now more easily switch between different Spotify accounts. The most recently active profile will always appear in the top right corner of a screen. Say you share a TV in the living room and you want to stop listening to your mom’s playlist – now you can select the profile image of a paired account when you want to switch.

In the past, if you listened to music on your TV through the app, you could only see the next song that was lined up. But now you can use the “Playback Queue” tool to select what you want to play next. Also, you can now turn on dark mode on your TV, meaning you can stream a curated playlist during your next house party and minimize the visuals and screen brightness.

The new Spotify TV app updates are available to free and premium subscribers using the app on supported Smart TVs, gaming consoles and media streaming devices. The effort to revitalize the user experience on the big screen isn’t a huge shocker. Spotify rolled out similar updates for its Desktop app earlier this year in an effort to make streaming more enjoyable when you’re not using your phone.

This article originally appeared on Engadget at https://www.engadget.com/spotify-for-tvs-gets-a-new-look-and-features-to-match-the-mobile-app-195050535.html?src=rss 

Amazon reportedly plans to remove the Android core from future Fire TVs, smart displays and other devices. Writing for Lowpass, veteran journalist Janko Roettgers cites “multiple sources with knowledge of these plans” along with job listings and other materials supporting Amazon’s alleged move to a new Linux-based operating system. The OS supposedly uses the internal codename “Vega” and runs on devices like Fire TV Sticks, TVs, and other connected devices.

Roettgers says Amazon has been working on the idea for years, with the company supposedly floating the idea to chipmakers as far back as 2017. “Hundreds of people” in Amazon’s Device OS group have reportedly worked on the software, including Zibi Braniecki, a former Mozilla engineer who joined Amazon in 2022. Earlier this year, he allegedly posted on LinkedIn (a comment that appears to have since been deleted) that he was “working on a next generation Operating System for Smart Home, Automotive, and other Amazon Devices product lines.”

The report doesn’t list a release timeline but mentions that “most of the OS development is already done.” Amazon is now apparently focusing on readying an SDK and planning perks to convince developers to invest their time and money.

The alleged move would allow Amazon to cut ties with Google’s Android Open Source Project, which lies at the heart of the current Fire OS. That dependence has led to Amazon’s software falling several generations behind the most recent Android versions. For example, the current-gen Fire TV software is based on Android 9, Google’s big software update from five years ago. Switching to Vega could also let Amazon cut underlying bloat from its OS; Android contains code enabling it to work on many different devices stretching far beyond the relatively minimal needs of Amazon’s smart-home gear.

Lowpass claims the Linux-based Vega would use React Native for app development. The Meta-created framework uses a single codebase for iOS and Android, and Amazon likely hopes that will help with cross-platform development. The report doesn’t specifically address whether the OS will eventually run on Fire tablets, which become much more versatile when users sideload the Google Play Store to run a wider variety of Android apps.

Roettgers’s report says Amazon plans to eventually make a clean break with Android on all new devices. The company allegedly designed Vega to run on systems as diverse as car infotainment systems and “other future hardware products.”

This article originally appeared on Engadget at https://www.engadget.com/amazon-reportedly-plans-to-dump-android-for-a-homemade-fire-os-replacement-200144781.html?src=rss 

Researchers at Cornell University have developed a wearable with batlike sonar that could improve upper-body tracking in virtual reality and other applications. The Cornell team fitted a generic pair of eyeglasses with a tiny sonar system, demonstrating how acoustic signals can be used instead of cameras to capture the body’s movement. 

Not only would sonar be more efficient in terms of battery consumption, the team told the Cornell Chronicle, but it would also do away with the privacy risks that come with headsets’ externally facing cameras. The system, dubbed PoseSonic, uses two pairs of microphones and speakers to send and receive acoustic signals, according to a recently published paper. With help from their deep learning model, it can then estimate 3D poses at nine different points — the shoulders, elbows, wrists, hips and nose — as these signals bounce off the upper body.

The team tested it both in the lab and “semi-in-the-wild,” and found it wasn’t negatively affected by environmental noise in any significant way. With this technique, “we use less instrumentation on the body, which is more practical, and battery performance is significantly better for everyday use,” senior author Cheng Zhang told the Cornell Chronicle.

In addition to its potential use in augmented and virtual reality, the researchers say sonar could make for better health tracking by capturing more detailed information on the body’s movements. They’ve only got the upper-body covered at the moment, though — VR legs continue to elude us.

This article originally appeared on Engadget at https://www.engadget.com/these-sonar-equipped-glasses-could-pave-the-way-for-better-vr-body-tracking-201958516.html?src=rss