Lifeboat Foundation

Genome editing is a powerful breeding technique that introduces mutations into specific gene sequences in genomes. For genome editing in higher plants, nucleotides for artificial nuclease (e.g. TALEN or CRISPR-Cas9) are transiently or stably introduced into the plant cells. After the introduction of mutations by artificial nucleases, it is necessary to select lines that do not contain the foreign nucleotides to overcome GMO regulation; however, there is still no widely legally authorized and approved method for detecting foreign genes in genome-edited crops. Recently, k-mer analysis based on next-generation sequencing (NGS) was proposed as a new method for detecting foreign DNA in genome-edited agricultural products. Compared to conventional methods, such as PCR and Southern hybridization, in principle, this method can detect short DNA fragments with high accuracy.

Researchers have long thought that rewards like food or money encourage learning in the brain by causing the release of the “feel-good” hormone dopamine, known to reinforce storage of new information. Now, a new study in rodents describes how learning still occurs in the absence of an immediate incentive.

Led by researchers at NYU Grossman School of Medicine, the study explored the relationship between dopamine and the brain chemical acetylcholine, also known to play a role in learning and memory. Past research had shown that these two hormones compete with one another, so that a boost in one causes a decline in the other. Rewards were thought to promote learning by simultaneously triggering an increase in dopamine and a decrease in acetylcholine.

This sudden hormone imbalance is believed to open a window of opportunity for brain cells to adjust to new circumstances and form memories for later use. Known as neuroplasticity, this process is a major feature of learning as well as recovery after injury. However, the question had remained whether food and other external rewards are the only drivers for this memory system, or whether our brains instead are able to create the same conditions that are favorable to learning without outside help.

That’s how Andrew Feldman, CEO of Silicon Valley AI computer maker Cerebras, begins his introduction to his company’s latest achievement: An AI supercomputer capable of 2 billion billion operations per second (2 exaflops). The system, called Condor Galaxy 1, is on track to double in size within 12 weeks. In early 2024, it will be joined by two more systems of double that size. The Silicon Valley company plans to keep adding Condor Galaxy installations next year until it is running a network of nine supercomputers capable of 36 exaflops in total.

If large-language models and other generative AI are eating the world, Cerebras’s plan is to help them digest it. And the Sunnyvale, Calif., company is not alone. Other makers of AI-focused computers are building massive systems around either their own specialized processors or Nvidia’s latest GPU, the H100. While it’s difficult to judge the size and capabilities of most of these systems, Feldman claims Condor Galaxy 1 is already among the largest.

Condor Galaxy 1—assembled and started up in just 10 days—is made up of 32 Cerebras CS-2 computers and is set to expand to 64. The next two systems, to be built in Austin, Texas, and Ashville, N.C., will also house 64 CS-2s each.

It’s been rumored for several months now that Apple will be using a new 3 nm manufacturing process from Taiwan Semiconductor (TSMC) for its next-generation chips, including M3 series processors for Macs and the A17 Bionic for some next-gen iPhones. But new reporting from The Information illuminates some of the favorable terms that Apple has secured to keep its costs down: Apple places huge chip orders worth billions of dollars, and in return, TSMC eats the cost of defective processor dies.

At a very high level, chip companies use large silicon wafers to create multiple chips at once, and the wafer is then sliced into many individual processor dies. It’s normal, especially early in the life of an all-new manufacturing process, for many of those dies to end up with defects—either they don’t work at all, or they don’t perform to the specifications of the company that ordered them.

Turtles migrate thousands of miles out in the open ocean, charting epic courses in search of food, mates, and nesting grounds. Exactly how they find where they’re going has long puzzled scientists who suspected magnetic fields were involved, but were unsure of the exact mechanism through which turtles were sensing it.

We’ve since learned that turtles appear to recognize magnetic signatures of locations, such as the beach on which they hatched where females will later return to lay their own eggs. We know the magnetosphere is in constant flux, and turtle nesting sites have been found to shift in tandem, so how is it that they’re able to make sense of this invisible force?

Some answers to this question were revealed in a study that looked at the way snapping turtles can tell north from south, in a phenomenon known as spontaneous magnetic alignment. It was once thought to be a rare trait in the animal kingdom, but as Professor John Phillips from the Department of Biological Sciences at Virginia Tech told IFLScience, this is no longer the case.

It’s one of the mysteries of nature: How does the axolotl, a small salamander, boast a superhero-like ability to regrow nearly any part of its body? For years, scientists have studied the amazing regenerative properties of the axolotl to inform wound healing in humans.

Now, Stanford Medicine researchers have made a leap forward in understanding what sets the axolotl apart from other animals. Axolotls, they discovered, have an ultra-sensitive version of mTOR, a molecule that acts as an on-off switch for protein production. And, like survivalists who fill their basements with non-perishable food for hard times, axolotl cells stockpile messenger RNA molecules, which contain genetic instructions for producing proteins. The combination of an easily activated mTOR molecule and a repository of ready-to-use mRNAs means that after an injury, axolotl cells can quickly produce the proteins needed for tissue regeneration.

The new findings were published July 26 in Nature.

Canadian architecture studio Atelier l’Abri has built a series of A-Frame buildings for the Farouche Tremblant agrotourism site in Québec’s Mon-Tremblant National Park, which were designed to “recede in the landscape”.

Intending to celebrate and showcase the surrounding untamed woodlands, Atelier l’Abri created a cafe, farm and four rental micro–cabins that act as a basecamp for visitors wanting to explore the nearby Devil’s River and its valley.

Sitting among the wild terrain, the four small rental cabins have steep-pitched roofs clad in cedar shingles that extend to the ground to form sloping walls.

Researchers revisit a neglected decay mode with implications for fundamental physics and for dating some of the oldest rocks on Earth and in the Solar System.

With a half-life of 1.25 billion years, potassium-40 does not decay often, but its decays have a big impact. As a relatively common isotope (0.012% of all potassium) of a very common metal (2.4% by mass of Earth’s crust), potassium-40 is one of the primary sources of radioactivity we encounter in daily life. Its decays are the primary source of argon-40, which makes up almost 1% of the atmosphere, and the copious amount of heat released from these decays threw off early estimates of the age of Earth made by Lord Kelvin. Potassium-40 is largely responsible for the meager radioactivity in our food (such as bananas), and it is a significant source of noise in some highly sensitive particle physics detectors. This isotope and its decay products are also useful tools in dating rocks and geological processes that go back to the earliest parts of Earth history. And yet some long-standing uncertainty surrounds these well-studied decays.

Let’s look at some examples of this software-defined momentum at the edge. In manufacturing, AI enables weld quality detection in real time on factory floors, improving production yields. In agriculture, farmers can use AI-driven systems to move from focusing on entire crops to looking at individual plants in a field to determine where to fertilize, irrigate or weed. Healthcare is transforming at every level—from the granularity of tracking nerve structures for anesthesia during surgery to the scale and scope of securing patient privacy and data across healthcare networks. An intelligent, software-defined edge aids in delivering resilience for evolving business needs.

AI tools and platforms are now widely available, allowing businesses to harness their power to build solutions faster and gain a competitive edge. This accessibility is crucial for scaling their usefulness, as it shifts solutions from being built solely by data scientists and software engineers to being used by domain experts with less coding experience. With simplified AI model toolkits and an open development platform, these users can stitch together their own solutions and deploy them anywhere.

Let’s take the example of a quick service restaurant (QSR). QSRs could improve their operations by monitoring orders and ingredient levels, then dynamically resupplying their inventories. Lowering barriers to AI means businesses like a QSR can tap into automation and intelligent software solutions on any device, such as a point-of-service system, laptop or mobile device. Customers are happier, food waste is reduced and process efficiencies help QSRs maintain operations even in our current labor shortage.