Tech and Hosting News Round-Up

Tech and Hosting News Round-Up

Welcome to our latest round-up of news from the technology and hosting world. Here’s what we’ve discovered this week.

Is this the most advanced social distancing robot?

If there is one type of technology which has come to the fore during the global lockdown, it is the robot. They’re being used to help people in care homes stay in touch, take people on virtual tours of galleries and, in many places, to enforce social distancing.

The latest and perhaps the most advanced social distancing robot seen so far is the Lima 002, a bionic dog-like robot that wouldn’t look out of place in a Star Wars movie. Indeed, not only does it look dog-like, it even moves like a dog, a feature that makes it perfect for patrolling uneven public places like parks.

Though it is manufactured in the USA, it is being used to keep the residents of Singapore at safe social distances and employs a variety of different technologies to make its way around locations, identify people and detect how close they are. However, rather than barking at those too near, a softly spoken female voice politely asks them to keep two metres apart.

Is it a bird? Is it a plane? No, it’s Superdrone!

The newest NHS hero to get an outing in Scotland is a drone that will deliver medical supplies between a hospital in Oban and another, offshore, on the Isle of Mull. Hospitals like these that serve remote communities often find it difficult to procure the essential equipment their front-line staff need to work safely and can struggle to get vital medical provisions to their patients.

The drones can help by carrying equipment directly from one hospital to another, helping them to share resources and better cater for their staff and patients. Without the need to travel over difficult terrain, wait for the next ferry service or send out drivers, the drones offer a faster and safer solution than transporting by car or van.

Facebook chatbot – a foul-mouthed liar?

Facebook’s aim with its new chatbot, Blender, was to improve AI interaction so that users were given a more human to human-like experience. In many ways this was a success: by using 1.5 billion Reddit interactions as part of its training, it has managed to become a knowledgeable and emphatic chatbot with a degree of personality. And, with over 10 billion parameters in its modelling, it has a sophisticated repertoire of conversational skills.

That repertoire, however, is perhaps too sophisticated and its personality all too human because, during tests, the chatbot let fly the occasional expletive and was prone to making up facts now and again. It certainly wasn’t bad news for the people who tested it, half of whom preferred talking to Blender than to a genuine human.

However, while the current version of Blender is likely to be the most fun to have a conversation with, it is not one Facebook will let us meet. Instead, we’ll get to interact with a Blender that’s been taught social etiquette and which refrains from swearing and telling fibs.  

Wanted Hackers! $100K Reward

Microsoft is looking for professional hackers to find vulnerabilities in its Linux-based Azure Sphere, a platform used by companies which deploy IoT devices. The software giant is offering up to $100k for any individual or team that can identify any security weaknesses that real hackers could exploit.

The project will take place over 3 months, from June to August, and is only open to those researchers and security firms approved by Microsoft. To earn the reward, those involved must be able to execute code on either the Azure Secure World or Azure Pluton. The aim of the project, according to Microsoft, is to provide security researchers with the content, resources and tools needed to defend high-impact, cloud vulnerabilities. No doubt, there’ll be some fun loving criminals out there who will also rise to the challenge.

Samsung’s toxic messages

Hackers have come up with a sneaky way to infect post-2015 Samsung smartphones, just by sending them a specially coded text message. The exploit, which affects Samsung’s default messaging app, is so malicious it works without user activation: there’s no need for the phone’s owner to open or read the message or click on any link, it simply starts to function as soon as it’s arrived.

The malicious message enables code to be executed which allows cybercriminals to spy on the owner and carry out various other unwanted operations. Samsung has been quick to respond and has released patching updates to supported phones which fix the vulnerability. Users should install this as soon as possible. If your phone model is not supported, you can overcome the problem by using an alternative messaging app instead of the default one supplied by Samsung.

Visit our website for more news, blog posts, knowledge base articles and information on our wide range of hosting services.

Sharing

Leave your comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.