Lifeboat Foundation

Alexander Britkin is the CEO of NFWare.

Digital transformation is accelerating at an unprecedented rate—and the telecommunications industry must evolve rapidly to meet demands for connectivity and performance. Telecom operators need to keep pace with network infrastructure developments to remain competitive and continue meeting customers’ short-and long-term needs. The following six trends are shaping the future of network infrastructure for telecom operators.

In less than 50 years, the internet has experienced explosive growth, far beyond the expectations of its creators. The internet’s original architecture, built in the 1970s, assigned each connected device a unique 32-bit number—an IPv4 (internet protocol version 4) address—to exchange information with other online devices. The initial bank of available IPv4 addresses totaled around 4.2 billion, which seemed substantial at the time, but demand has already greatly exceeded supply.

‘Space Gal’ is about to take on a much more literal meaning.

Some sequences in the genome cause genes to be switched on or off. Until now, each of these gene switches, or so-called enhancers, was thought to have its own place on the DNA. Different enhancers are therefore separated from each other, even if they control the same gene, and switch it on in different parts of the body.

A recent study from the University of Bonn and the LMU Munich challenges this idea. The findings are also important because gene switches are thought to play a central role in evolution. The study has been published in the journal Science Advances.

The blueprint of plant and animal forms is encoded in their DNA. But only a small part of the genome—about two percent in mammals—contains genes, the instructions for making proteins. The rest largely controls when and where these genes are active: how many of their transcripts are produced, and thus how many proteins are made from these transcripts.

Good news for anyone with a hankerin’ for going back in time to kill their grandfather before he had kids: a physicist named Germain Tobar from the University of Queensland in Australia says go for it since time travel paradoxes aren’t real. So feel free to kill your grandpappy without fear of deleting your own existence.

He didn’t explicitly frame it that way, but he does think that time travel paradoxes are bullshit. Tobar’s work uses Einstein’s theory of general relativity as a foundation and then builds from there. He says that, according to his calculations, events can exist both in the past and in the future simultaneously, independent of one another. Space-time will adjust itself to avoid paradoxes, thus allowing you to cause whatever mayhem you want throughout time without creating contradictions.

If true, famous time travel stories like The Terminator and Back to the Future wouldn’t be possible. A Terminator sent to the past to kill John Connor would not be killing John Connor in the future, theoretically. It would only kill John Connor in the past and space-time would find some way to adjust to ensure that John Connor is still alive in the future to continue to be a pain in every robot’s shiny metal ass.

The delivery of nanomedicines using gas bubbles has shown itself to be a unique way of transporting cytotoxins to the lungs of cancer patients. The method enables precise and focused treatments, and the local action of the drugs also prevents a range of side-effects.

In today’s AI news, Nvidia has three anonymous customers, referred to as Customers A, B, and C, who have each spent over $10 Billion on the company’s AI chips so far this year.

However, Hassabis’ true breakthrough came just a month ago, when he and two colleagues from DeepMind won the Nobel Prize in Chemistry for their development of AlphaFold, an AI tool capable of predicting the structure of the 200 million known proteins. This achievement would have been nearly impossible without AI, and solidifies Hassabis’ belief that AI is set to become one of the main drivers of scientific progress in the coming years.

Hassabis — the son of a Greek-Cypriot father and a Singaporean mother — reflects on the early days of DeepMind, which he founded in 2010, when “nobody was working on AI.” Over time, machine learning techniques such as deep learning and reinforcement learning began to take shape, providing AI with a significant boost. In 2017, Google scientists introduced a new algorithmic architecture that enabled the development of AGI. “It took several years to figure out how to utilize that type of algorithm and then integrate it in hybrid systems like AlphaFold, which includes other components,” he explains.

“During our first years, we were working in a theoretical space. We focused on games and video games, which were never an end in themselves. It gave us a controlled environment in which to operate and ask questions. But my passion has always been to use AI to accelerate scientific understanding. We managed to scale up to solving a real-world problem, such as protein folding,” recalls the engineer and neuroscientist.

Frontier uses a network of thousands of nodes, each connected by more than 90 miles of cable, which allows them to collaborate on complex issues.

Imagine sitting down with an AI model for a spoken two-hour interview. A friendly voice guides you through a conversation that ranges from your childhood, your formative memories, and your career to your thoughts on immigration policy. Not long after, a virtual replica of you is able to embody your values and preferences with stunning accuracy.

That’s now possible, according to a new paper from a team including researchers from Stanford and Google DeepMind, which has been published on arXiv and has not yet been peer-reviewed.