Distributed Machine Learning (DML) on resource-constrained edge devices holds immense potential for real-world applications. However, achieving fast convergence in DML in these heterogeneous environments remains a significant challenge. Traditional frameworks like Bulk Synchronous Parallel and Asynchronous Stochastic Parallel rely on frequent, small updates that incur substantial communication overhead and hinder convergence speed. Furthermore, these frameworks often employ static dataset sizes, neglecting the heterogeneity of edge devices and potentially leading to straggler nodes that slow down the entire training process. The straggler nodes, i.e., edge devices that take significantly longer to process their assigned data chunk, hinder the overall training speed. To address these limitations, this paper proposes Hermes, a novel probabilistic framework for efficient DML on edge devices. This framework leverages a dynamic threshold based on recent test loss behavior to identify statistically significant improvements in the model's generalization capability, hence transmitting updates only when major improvements are detected, thereby significantly reducing communication overhead. Additionally, Hermes employs dynamic dataset allocation to optimize resource utilization and prevents performance degradation caused by straggler nodes. Our evaluations on a real-world heterogeneous resource-constrained environment demonstrate that Hermes achieves faster convergence compared to state-of-the-art methods, resulting in a remarkable $13.22$x reduction in training time and a $62.1\%$ decrease in communication overhead.

A new translation feature automatically translates any text in a design to over 100 different languages, and video footage can be matched up to the beat of a soundtrack without any manual edits. Canva said that it uses a mix of different AI models as the foundation for these features in addition to building its own systems. Finally, 953 new fonts and a host of non-workspace-specific tools are being added to the design platform that have been requested by its wider community. New editing features for layouts, layers, styles, and gradients are available to jazz up your designs, and alt text can now be generated for images directly within Canva.

The world of human artists will undoubtedly be shaken in the next months by this significant advancement for text-to-image AI models, for better or worse. And although I don't intend to bombard my email subscribers with AI image generator news, a lot of big updates are getting released every single day, I can't help but get amazed and share my excitement with you guys.

Scientists have built a shape-shifting robotic drone that transforms from a ground vehicle to a quadcopter, an advance that may lead to novel machines that can perform diverse functions under different conditions and self-heal after being damaged. The researchers from Virginia Tech in the US first developed a multifunctional material that could change shape, hold the shape, and return to the original configuration, and to do this over many cycles. "One of the challenges was to create a material that was soft enough to dramatically change shape, yet rigid enough to create adaptable machines that can perform different functions," Michael Bartlett, assistant professor in mechanical engineering, said in a statement. The scientists then turned to the Japanese art of kirigami, which involves making shapes out of paper by cutting, to create a structure that could be morphed. Then they developed an endoskeleton made of a low melting point alloy (LMPA) embedded inside a rubber skin.

Tesla car owner Draper Younce says his Tesla is a huge reason why he survived an attempted carjacking. Tesla has canceled a planned expansion of its Full Self-Driving beta software test ahead of a major update reportedly launching in April. Approximately 2,000 Tesla owners have been given early access to the expanded functionality, which allows a vehicle to control itself on city streets with human supervision and is listed as version 8.2 of the software. In early March, Tesla CEO Elon Musk said that a download button for version 8.3 would soon appear on the displays in certain cars, but he has apparently confirmed on Twitter that the company will be skipping the upgrade in lieu of a version 9.0 coming in April. Musk last week said that the upgrade would eliminate the need for the system to use radar and would rely solely on the eight cameras the cars are equipped with and artificial intelligence.

The new BBC micro:bit'mini-computer' given to school children has been upgraded with AI and machine learning support, a new speaker and a microphone. It was launched in 2016 as part of the BBC'Make it Digital' campaign and four years later over five million have been used by schools and children around the world. The project is no longer run by the BBC, it was taken over by the Micro Bit Educational Foundation, a non-profit group setup to make coding more accessible. As well as a new microphone and speaker, that could be used for anything from listening out for doorbell sounds to playing back voice recordings, the device will include a touch sensor that could count how often a fly lands on a pad. The new BBC micro:bit'mini-computer' given to school children has been upgraded with AI and machine learning support, a new speaker and a microphone The palm-sized micro:bit was launched in 2016 as part of the BBC'Make it Digital' campaign and four years later over five million have been used by schools and children around the world Other new sensors on the device including light, magnetism and temperature, to create a wider range of applications.

Tesla CEO Elon Musk has said the company's Autopilot will'begin to enable full self-driving features' this summer. The billionaire tweeted that a major update to the software coming in August, dubbed Tesla Version 9, will fix a number of issues with the Autopilot system. He said the upgrade will resolve problems with lane merging and enhance the software's self-driving capabilities. The news comes days after federal investigators determined that a Tesla SUV using the firm's autonomous driving system accelerated before it collided with a California freeway divider in March, killing its driver. Data shows the Model X SUV did not brake or try to steer around the barrier in the three seconds before the crash. The driver did not have his hands on the wheel.

Microsoft is planning to include more artificial intelligence capabilities inside Windows 10 soon. The software giant is unveiling a new AI platform, Windows ML, for developers today, that will be available in the next major Windows 10 update available this spring. Microsoft's new platform will enable all developers that create apps on Windows 10 to leverage existing pre-trained machine learning models in apps. Windows ML will enable developers to create more powerful apps for consumers running Windows 10. Developers will be able to import existing learning models from different AI platforms and run them locally on PCs and devices running Windows 10, speeding up real-time analysis of local data like images or video, or even improving background tasks like indexing files for quick search inside apps.

Microsoft has announced that its next major update to Windows 10 will include Windows ML, the company's new AI platform. Windows ML will let developers use machine learning systems running in their apps, which could result in faster artificial intelligence. "With the next major update to Windows 10, we begin to deliver the advances that have been built into our apps and services as part of the Windows 10 platform," the company said on its blog. "Every developer that builds apps on Windows 10 will be able to use AI to deliver more powerful and engaging experiences." Microsoft says that the inclusion of Windows ML in the next Windows 10 update will offer a number of benefits to developers.

The technology will support the "industry standard" ONNX format for machine learning models, so AI veterans won't necessarily have to reinvent the wheel to incorporate their work. Microsoft hasn't said when that Windows 10 update will arrive beyond this year, although programmers will get an early peek at it in the Visual Studio Preview 15.7 release. You should hear more about the feature both at its Windows Developer day on March 7th as well as at the Build conference in May. This isn't an unheralded event when Google is already including an AI platform (TensorFlow Lite) as part of Android Oreo, but it's relatively new in the PC world and could have a transformative effect. Microsoft is hoping that you'll see "more intelligence" across Windows devices, even if it's for simple tasks like auto-generated music playlists -- it won't have to be limited to developers with deep enough pockets to tackle AI on their own.