Author: chemistadmin

Google Gemini Can Summarize Your Emails in Gmail. Should You Use It?

[ad_1]

Artificial intelligence is now busy tackling some of the biggest problems to face humankind: Speeding up drug design, tackling cancer detection, and finding solutions to climate change. However, none of these issues are arguably as daunting as the task Google has set its Gemini AI bot on.

Specifically, the task of staying on top of your inbox. Gemini is now a part of Gmail on the web and on mobile devices, and as well as using it to find the right words in your emails, you can also get it to summarize long emails and threads for you.

Here I’ll show you how these summary tools work and what else Gemini can do for you—and report on just how reliable it is at the moment. One caveat though: For now, Gemini in Gmail is only available if you or your employer are paying for Google One AI Premium ($20 a month), or for a Google Workspace account.

Get Gemini Summaries in Gmail

Gemini can summarize single emails, or lots of them.David Nield

There are a few ways to get Gemini summaries in Gmail, if the feature is enabled for your account. Most of them can be accessed through the Gemini logo, which is a distinctive black star shape. On the web, click the Gemini button in the top right corner of Gmail to bring up the side panel. There, you can see summaries for your inbox as a whole, or for the particular thread you have open.

In Gmail for Android and iOS, the Gemini button shows up in the top right corner if you’re looking at a list of emails, or in the center at the top if you’re viewing a particular thread. On mobile, there’s also a specific Summarize this email button that appears when you’re looking at a single email or a single thread of emails.

That Summarize this email button is the easiest way to get started, but you can also tell Gemini to “summarize today’s emails,” “summarize this week’s emails,” “summarize my unread emails,” or “summarize the emails I got last month”—anything along those lines. After Gemini spends a few moments thinking, you’ll get a response on screen, together with follow-up questions you might want to ask. (You can request a longer summary, for instance.)

The results will be presented as a series of bullet points, with Sources underneath: Click or tap on these sources to see the individual emails the information was pulled from. Using the icons alongside the responses, you’re also able to copy the text elsewhere, give thumbs up or thumbs down feedback on the Gemini response, or clear the AI chat history.

Ask Gemini Other Questions in Gmail

There’s more to Gemini than summaries.David Nield

I’m mostly focusing on the summary capabilities of Gemini in Gmail here, but there are plenty of other commands you can explore. In fact, you can ask Gemini just about any question you like about what’s in your inbox, and it will at least attempt to provide a response—scouring through the gigabytes of data in your emails looking for answers.

[ad_2]

Source link

December 9, 2024
People who are good at reading have different brains

[ad_1]

The number of people who read for fun appears to be steadily dropping. Fifty percent of UK adults say they don’t read regularly (down from 58% in 2015) and almost one in four young people aged 16-24 say they’ve never been readers, according to research by The Reading Agency.

But what are the implications? Will people’s preference for video over text affect our brains or our evolution as a species? What kind of brain structure do good readers actually have? My new study, published in Neuroimage, has found out.

I analysed open-source data from more than 1,000 participants to discover that readers of varying abilities had distinct traits in brain anatomy.

The structure of two regions in the left hemisphere, which are crucial for language, were different in people who were good at reading.

One was the anterior part of the temporal lobe. The left temporal pole helps associate and categorise different types of meaningful information. To assemble the meaning of a word such as leg, this brain region associates the visual, sensory and motor information conveying how legs look, feel and move.

The other was Heschl’s gyrus, a fold on the upper temporal lobe which hosts the auditory cortex (the cortex is the outermost layer of the brain). Better reading ability was linked to a larger anterior part of the temporal lobe in the left hemisphere compared to the right. It makes sense that having a larger brain area dedicated to meaning makes it easier to understand words and, therefore, to read.

What might seem less intuitive is that the auditory cortex would be related to reading. Isn’t reading mainly a visual skill? Not only. To pair letters with speech sounds, we first need to be aware of the sounds of the language. This phonological awareness is a well-established precursor to children’s reading development.

A thinner left Heschl’s gyrus has previously been related to dyslexia, which involves severe reading difficulties. My research shows that this variation in cortical thickness does not draw a simple dividing line between people with or without dyslexia. Instead, it spans the larger population, in which a thicker auditory cortex correlates with more adept reading.

Why size matters

Is thicker always better? When it comes to cortical structure, no, not necessarily. We know the auditory cortex has more myelin in the left hemisphere of most people. Myelin is a fatty substance that acts as an insulator for nerve fibres. It increases neural communication speed and can also insulate columns of brain cells from each other. Neural columns are believed to function as small processing units.

Their increased isolation and rapid communication in the left hemisphere can be thought to enable the fast, categorical processing necessary for language. We need to know if a speaker uses the category d or t when saying dear or tear rather than detecting the exact point where the vocal folds start vibrating.

According to the “balloon model” of cortical growth, the larger amount of myelin squeezes out left-hemispheric cortical areas, making them flatter but more extended. So while the left auditory cortex may be thicker in good readers, it is still thinner (but much more extended) than the corresponding right cortex.

This hypothesis was corroborated in the recent research. The left hemisphere had generally larger but thinner cortical areas with a higher degree of myelin.

So is thinner better, then? Again, the answer is no, not necessarily. Complex abilities that require integrating information tend to benefit from a thicker cortex. The anterior temporal lobe with its complex way of integrating information is indeed the thickest structure of all cortical areas. An underlying mechanism might be the existence of more overlapping, interacting neurons which process information more holistically.

Phonology is a highly complex skill, where different sound and motor features are integrated into speech sounds. It appears to correlate with a thicker cortex in an area near the left Heschl’s gyrus. While it is unclear to what extent phonology is processed in Heschl’s gyrus, the fact that phoneticians often have multiple left Heschl’s gyri suggests it is linked to speech sounds.

The temporal lobe is involved in reading.
Shutterstock

Clearly, brain structure can tell us a lot about reading skills. Importantly, though, the brain is malleable — it changes when we learn a new skill or practice an already acquired one.

For instance, young adults who studied language intensively increased their cortical thickness in language areas. Similarly, reading is likely to shape the structure of the left Heschl’s gyrus and temporal pole. So, if you want to keep your Heschl’s thick and thriving, pick up a good book and start reading.

Finally, it’s worth considering what might happen to us as a species if skills like reading become less prioritised. Our capacity to interpret the world around us and understand the minds of others would surely diminish. In other words, that cosy moment with a book in your armchair isn’t just personal – it’s a service to humanity.

[ad_2]

Source link

December 9, 2024
Believing in Santa Claus doesn’t make children act nicer at Christmas

[ad_1]

Santa Claus alone is not enough for a happy Christmas

Tristan Fewings/Getty Images for Hamleys

He’s making a list, he’s checking it twice, but Santa’s festive surveillance seemingly does nothing to improve children’s behaviour. Instead, it may be that wider Christmas rituals, like putting up a tree and going carolling, can prompt children to be a bit nicer – a finding that may help us better understand how religion influences behaviour.

“The question was, does belief in Santa Claus influence how children behave?” says Rohan Kapitány at Durham University in the UK. “Does this belief,…

[ad_2]

Source link

December 9, 2024
The Sticky Dilemmas of Pornhub’s Next Chapter

[ad_1]

Videos of minors. Illegal data collection. Lack of oversight. Lawsuits. Problems have dogged the popular porn site for years. Is its promise of transparency enough for a reset?

[ad_2]

Source link

December 9, 2024
Meet the Plant Hacker Creating Flowers Never Seen (or Smelled) Before

[ad_1]

The money he earned doing that was enough to put Cocioba through the first couple of years of a biology degree at Stony Brook University. He completed a stint with a neglected plant biology group that taught him to experiment on a shoestring budget. “We were using toothpicks and yogurt cups to do petri dishes and all of that,” he says. But financial difficulties meant he had to drop out. Before he left, one of his labmates handed him a tube of agrobacterium—a microbe commonly used to engineer new attributes into plants.

A Petunia bioengineered by Sebastian Cocioba, a plant biotechnology researcher who works out of his home laboratory in Huntington, New York on October 30, 2024.Lanna Apisukh

A shelf of bio engineered plants under grow lights in Sebastian Cocioba’s home on October 30, 2024. The plant biotechnology researcher built a laboratory inside his home where he works out of in Huntington, New York.Lanna Apisukh

Test tubes of Petunias under a grow light in Huntington, New York on October 30, 2024. The flowers were bioengineered by Sebastian Cocioba, a plant biotechnology researcher who works out of his home laboratory.Lanna Apisukh

Cocioba set about transforming his hallway nook into a makeshift lab. He realized that he could buy cheap equipment in fire sales from labs that were shutting down and sell them on for a markup. “That gave me a little bit of an income stream,” he says. Later he learned to 3D-print relatively simple pieces of equipment that are sold at extreme markups. A light box used to visualize DNA, for example, could be cobbled together with some cheap LEDs, a piece of glass, and a light switch. The same device would retail to laboratories for hundreds of dollars. “I have this 3D printer, and it’s been the most enabling technology for me,” Cocioba says.

All of this tinkering was in aid of Cocioba’s main mission: to become a flower designer. “Imagine being the Willy Wonka of flowers, without the sexism, racism, and strange little slaves,” he says. In the US, genetically modified flower work is covered by the lowest biosafety rating, so it doesn’t subject Cocioba or his lab to onerous regulations. Doing gene-editing as an amateur in the UK or EU would be impossible, he says.

Cocioba set himself up as a self-described “pipette for hire”—working for startups to develop scientific proof-of-concepts. In the run-up to the 2020 Tokyo Olympics, the plant biologist Elizabeth Hénaff asked Cocioba for help with a project she was working on: designing a morning glory flower with the Games’ blue-and-white checkerboard pattern. It just so happened that a checkerboard flower already existed in nature—the snake’s head fritillary. Cocioba wondered if he could import some of the genes from that plant into a morning glory. Unfortunately it turned out that the snake’s head fritillary had one of the largest genomes on the planet and had never been sequenced. With the Olympics looming, the project fell apart. “It ended in heartbreak, of course, because we couldn’t execute on it.”

A close-up view of Petunia tissue culture grown by Sebastian Cocioba, a plant biotechnology researcher based in Huntington, New York on October 30, 2024.Lanna Apisukh

Test tubes of frozen DNA and plant enzymes inside the home laboratory of Sebastian Cocioba, a plant biotechnology researcher based in Huntington, New York on October 30, 2024.Lanna Apisukh

As Cocioba moved deeper into the world of synthetic biology, he started to shift his focus slightly—away from just creating new kinds of plants and toward opening up the tools of science itself. Now he documents his experiments on an online notebook that’s free for anyone to use. He also started selling some of the plasmids—small circles of plant DNA—that he uses to transform flowers.

“We’re at the golden age of biotech for sure,” he says. Access is greater, and the research community is more open than ever before. Cocioba is trying to recreate something like the 19th-century boom of amateur plant breeders—where hobbyist scientists shared their materials partly just for the thrill of creating new plant varieties. “You don’t have to be a professional scientist to do science,” Cocioba says.

Alongside this work, Cocioba is also a project scientist at the California-based startup Senseory Plants. The company wants to engineer indoor plants to produce unique scents—a biological alternative to candles or incense sticks. One idea he’s playing with is engineering a plant to smell like old books, olfactorily transforming a room into an ancient library. The startup is exploring a whole smellscape of evocative scents, Cocioba says, in part designed in his home laboratory. “I really, really, love what they’re doing.”

This article appears in the January/February 2025 issue of WIRED UK magazine.

[ad_2]

Source link

December 9, 2024
Advancing flood risk analysis: transitioning from bathtub modelling

[ad_1]

Researchers from the University of California, Irvine, and the University of Bristol are emphasising the need to abandon outdated methods of flood hazard mapping, specifically a technique known as “bathtub modelling.”

By adopting more advanced methods, scientists aim to improve the accuracy and reliability of flood risk assessments, which are essential for communities, governments, and industries.

The importance of accurate flood modelling

Accurate flood modelling is crucial for managing the growing risks posed by climate change. Inaccurate models not only undermine public trust but can also hinder progress in several key areas.

Communities depend on reliable projections to understand their vulnerabilities and take action to protect themselves. Policymakers and insurers rely on accurate data to allocate resources and develop infrastructure that can withstand future flood events.

Furthermore, trustworthy models are essential for implementing equitable and effective flood mitigation strategies. Without accurate assessments, resources may be distributed inequitably, leaving vulnerable populations at greater risk.

Economic stability also hinges on accurate modelling, as businesses in flood-prone areas require precise risk assessments to make sound operational and investment decisions.

By transitioning to advanced modelling techniques, scientists can provide a more accurate and comprehensive picture of flood risks.

This, in turn, empowers communities, governments, and industries to take meaningful and evidence-based actions to mitigate the impacts of flooding.

The problem with bathtub modelling

Bathtub modelling, a method that assumes floodwaters spread evenly across affected areas like water filling a bathtub, has long been used to visualise potential flood impacts.

While straightforward and visually intuitive, this method often fails to represent the complexities of real-world flooding.

This technique does not account for key factors that influence flood behaviour, including the role of storm drains, levees, and pumping systems that manage water flow.

Additionally, it overlooks tidal amplification, the impact of rising groundwater, and the friction and dynamics that shape flood events.

As a result, bathtub models can both overestimate and underestimate flood risks, leading to unreliable hazard maps.

The researchers point out that studies using bathtub modelling often produce critical success index (CSI) scores below 0.5.

This score, which measures the accuracy of flood extent predictions, indicates that bathtub models perform worse than a random classification.

The authors argue that models need a CSI of at least 0.65 to be locally relevant and useful for practical applications.

Misleading projections and their impact

The limitations of bathtub modelling are compounded by the way its results are sometimes presented to the public.

Dramatic visualisations of flooded cities, often based on oversimplified data, can mislead policymakers and the public about the true risks of flooding.

While these images may attract attention, they fail to provide actionable insights, potentially leading to poor decision-making.

For communities in flood-prone areas, inaccurate models can erode trust. If projections do not align with lived experiences, residents may be less likely to prepare for future risks.

Conversely, exaggerated forecasts can create unnecessary panic or lead to inefficient allocation of resources for flood defences. These missteps emphasise the need for more precise flood risk assessments to guide policy and planning effectively.

A call for change

The researchers behind this study stress that the climate science community must move away from outdated approaches like bathtub modelling.

While these methods have been widely used for their simplicity, their inaccuracies make them unsuitable for the growing challenges posed by climate change and rising sea levels.

Improved modelling techniques are needed to produce trustworthy flood hazard maps that can guide mitigation strategies, inform policymaking, and empower communities.

Without accurate and reliable models, efforts to adapt to increasing flood risks may fall short, leaving people and assets vulnerable to devastating impacts.

As the frequency and severity of flooding events continue to rise, ensuring that projections are both scientifically robust and practically relevant is more important than ever.

By adopting more advanced methods, scientists can provide the tools necessary to build resilience in the face of a changing climate.

[ad_2]

Source link

December 9, 2024
OpenAI’s “12 days of shipmas” tell us a lot about the AI arms race

[ad_1]

The Chat

Every week, I talk to one of MIT Technology Review’s journalists to go behind the scenes of a story they are working on. This week, I hit up Amanda Silverman, our features and investigations editor, about our big story on the way the war in Ukraine is reshaping the tech sector in eastern Europe.

Mat: Amanda, we published a story this week from Peter Guest that’s about the ways civilian tech is being repurposed for the war in Ukraine. I could be wrong, but ultimately I think it showed how warfare has truly changed thanks to inexpensive, easily-built tech products. Is that right?

Amanda: I think that’s pretty spot on. Though maybe it’s more accurate to say, less expensive, more-easily-built tech products. It’s all relative, right? Like, the retrofitted consumer drones that have been so prevalent in Ukraine over the past few years are vastly cheaper than traditional weapons systems, and what we’re seeing now is that lots of other tech that was initially developed for civilian purposes—like, Pete reported on a type of scooter—are being sent to the front. And again, these are much, much cheaper than traditional weaponry. And they can be developed and shipped out really quickly.

The other thing Pete found was that this tech is being quickly reworked to respond to battlefield feedback—like that scooter has been customized to carry NATO standard-sized bullet boxes. I can’t imagine that happening in the old way of doing things.

Mat: It’s move fast and (hope not to) break things, but for war…. There is also this other, much scarier idea in there, which is that the war is changing, maybe has changed, Eastern Europe’s tech sector. What did Pete find is happening there?

Amanda: So a lot of the countries neighboring Ukraine are understandably pretty freaked out by what happened there and how the country had to turn on a dime to respond to the full-scale invasion by Russia. At the same time, Pete found that a lot of people in these countries, particularly in Latvia and particularly leading tech startups, have been inspired by how Ukrainians mobilized for the war and they’re trying to sort of get ahead of the potential enemy and get ready for a conflict within their borders. It’s not all scary, to be clear. It’s arguably somewhat thrilling to see all this innovation happening so quickly and to have some of the more burdensome red tape removed.

Mat: Okay so Russia’s neighbors are freaked out, as you say, understandably. Did anything about this story freak you out?

Amanda: Yeah, it’s impossible to ignore that there is a huge, scary risk here, too: as these companies develop new tech for war, they have an unprecedented opportunity to test it out in Ukraine without going through the traditional development and procurement process—which can be slow and laborious, sure, but also includes a lot of important testing, checks and balances, and more to prevent fraud and lots of other abuses and dangers. Like, Pete nods to how Clearview AI was deploying its tech to identify Russian war dead, which is scary in and of itself and also may violate the Geneva Conventions.

[ad_2]

Source link

December 9, 2024
Chips linked with light could train AI faster while using less energy
[ad_1]
The prototype of an IBM optics module for connecting chips with fibre optics

Ryan Lavine for IBM

An optical fibre technology can help chips communicate with each other at the speed of light, enabling them to transmit 80 times as much information as they could using traditional electrical connections. That could significantly speed up the training times required for large artificial intelligence models – from months to weeks – while also reducing the energy and emissions costs for data centres.

Most advanced computer chips still communicate using electrical signals carried over copper wires. But as the tech industry races to train large AI models – a process that requires networks of AI superchips to transfer huge amounts of data – companies are eager to link chips using the light-speed communication of fibre optics.

This technology isn’t new: the internet already relies on undersea fibre-optic cables stretching thousands of kilometres between continents. In order to transmit data between fingernail-size chips, however, companies must connect as many hair-thin optical fibres as possible to the edge of each chip.

“As we all know, the best communication technology is fibre optics, and that’s why fibre optics is used everywhere else for long-distance communication,” said Mukesh Khare at IBM Research during a press briefing previewing the technology. “This co-packaged optics innovation is basically bringing the power of fibre optics on the chip itself.”

Khare and his colleagues have developed an optics module that would enable chipmakers to add six times as many optical fibres to the edge of a chip, compared to current technologies. The module uses a structure called an optical waveguide to connect as many as 51 optical fibres per millimetre. It also prevents light signals from one fibre from interfering with its neighbours.

“What IBM has really done here is use all of its materials and packaging technology – the history of leadership in that – to really break through how you do high-density fibre optics by using waveguides,” says Dan Hutcheson at TechInsights, a semiconductor tech research firm headquartered in Canada. “To me, that was the big breakthrough when I saw it.”

The resulting boost in communication between chips could enable AI developers to train a large language model within three weeks instead of three months. Switching from electrical wires to optical fibres for chip communication could also mean a fivefold reduction in the energy cost of training such AI models.

IBM has already put the optical module through stress tests that included high humidity and temperatures ranging from -40°C (-40°F) to 125°C (257°F). Hutcheson expects that major semiconductor manufacturing companies may be interested in licensing the technology.

“We’re really in the early days of all this, but it’s the hottest area in semiconductor technology right now in terms of high-performance computing and AI technology,” he says.
Topics:

artificial intelligence/

computing
[ad_2]

Source link
December 9, 2024
Rentering spacecraft could hurt Earth’s atmosphere

[ad_1]

Changing chemistry

None of the researchers aboard the business jet turned scientific laboratory that took off from Easter Island in September got to see the moment when Cluster Salsa burst into a fireball above the deep, dark waters of the Pacific Ocean. Against the bright daylight, the fleeting explosion appeared about as vivid as a midday full moon. The windows of the plane, however, were covered with dark fabric (to prevent light reflected from inside to skew the measurements), allowing only the camera lenses to peek out, says Jiří Šilha, CEO of Slovakia-based Astros Solutions, a space situational awareness company developing new techniques for space debris monitoring, which coordinated the observation campaign.

“We were about 300 kilometers [186 miles] away when it happened, far enough to avoid being hit by any remaining debris,” Šilha says. “It’s all very quick. The object reenters at a very high velocity, some 11 kilometers [seven miles] per second, and disintegrates 80 to 60 kilometers above Earth.”

ESA

The instruments collected measurements of the disintegration in the visible and near-infrared part of the light spectrum, including observations with special filters for detecting chemical elements including aluminum, titanium, and sodium. The data will help scientists reconstruct the satellite breakup process, working out the altitudes at which the incineration takes place, the temperatures at which it occurs, and the nature and quantity of the chemical compounds it releases.

The dusty leftovers of Cluster Salsa have by now begun their leisurely drift through the mesosphere and stratosphere—the atmospheric layers stretching at altitudes from 31 to 53 miles and 12 to 31 miles, respectively. Throughout their decades-long descent, these ash particles will interact with atmospheric gases, causing mischief, says Connor Barker, a researcher in atmospheric chemical modeling at University College London and author of a satellite air pollution inventory published in early October in the journal Scientific Data.

Satellite bodies and rocket stages are mostly made of aluminum, which burns into aluminum oxide, or alumina—a white, powdery substance known to contribute to ozone depletion. Alumina also reflects sunlight, which means it could alter the temperature of those higher atmospheric layers.

“In our simulations, we start to see a warming over time of the upper layers of the atmosphere that has several knock-on effects for atmospheric composition,” Barker says.

For example, some models suggest the warming could add moisture to the stratosphere. This could deplete the ozone layer and could cause further warming, which in turn would cause additional ozone depletion.

[ad_2]

Source link

December 9, 2024
From Data to Insight: Mastering Visualization for Quality Control and Chemometrics

[ad_1]

The aim of this workshop is to encourage participants to discuss their experiences and challenges regarding data visualization in the field of chemometrics. The event will build a platform for learning about effective data visualization, professional networking, and scientific knowledge exchange.

Event Details

[ad_2]

Source link

December 9, 2024