Blog

What the EU’s tough AI law means for research and ChatGPT

[ad_1]

Representatives of EU member governments approved the EU AI Act this month.Credit: Jonathan Raa/NurPhoto via Getty

European Union countries are poised to adopt the world’s first comprehensive set of laws to regulate artificial intelligence (AI). The EU AI Act puts its toughest rules on the riskiest AI models, and is designed to ensure that AI systems are safe and respect fundamental rights and EU values.

“The act is enormously consequential, in terms of shaping how we think about AI regulation and setting a precedent,” says Rishi Bommasani, who researches the societal impact of AI at Stanford University in California.

The legislation comes as AI develops apace. This year is expected to see the launch of new versions of generative AI models — such as GPT, which powers ChatGPT, developed by OpenAI in San Francisco, California — and existing systems are being used in scams and to propagate misinformation. China already uses a patchwork of laws to guide commercial use of AI, and US regulation is under way. Last October, President Joe Biden signed the nation’s first AI executive order, requiring federal agencies to take action to manage the risks of AI.

EU nations’ governments approved the legislation on 2 February, and the law now needs final sign-off from the European Parliament, one of the EU’s three legislative branches; this is expected to happen in April. If the text remains unchanged, as policy watchers expect, the law will enter into force in 2026.

Some researchers have welcomed the act for its potential to encourage open science, whereas others worry that it could stifle innovation. Nature examines how the law will affect research.

What is the EU’s approach?

The EU has chosen to regulate AI models on the basis of their potential risk, by applying stricter rules to riskier applications and outlining separate regulations for general-purpose AI models, such as GPT, which have broad and unpredictable uses.

The law bans AI systems that carry ‘unacceptable risk’, for example those that use biometric data to infer sensitive characteristics, such as people’s sexual orientation. High-risk applications, such as using AI in hiring and law enforcement, must fulfil certain obligations; for example, developers must show that their models are safe, transparent and explainable to users, and that they adhere to privacy regulations anddo not discriminate. For lower-risk AI tools, developers will still have to tell users when they are interacting with AI-generated content. The law applies to models operating in the EU and any firm that violates the rules risks a fine of up to 7% of its annual global profits.

“I think it’s a good approach,” says Dirk Hovy, a computer scientist at Bocconi University in Milan, Italy. AI has quickly become powerful and ubiquitous, he says. “Putting a framework up to guide its use and development makes absolute sense.”

Some don’t think the laws go far enough, leaving “gaping” exemptions for military and national-security purposes, as well as loopholes for AI use in law enforcement and migration, says Kilian Vieth-Ditlmann, a political scientist at AlgorithmWatch, a Berlin-based non-profit organization that studies the effects of automation on society.

How much will it affect researchers?

In theory, very little. Last year, the European Parliament added a clause to the draft act that would exempt AI models developed purely for research, development or prototyping. The EU has worked hard to make sure that the act doesn’t affect research negatively, says Joanna Bryson, who studies AI and its regulation at the Hertie School in Berlin. “They really don’t want to cut off innovation, so I’d be astounded if this is going to be a problem.”

The European Parliament must give the final green light to the law. A vote is expected in April.Credit: Jean-Francois Badias/AP via Alamy

But the act is still likely to have an effect, by making researchers think about transparency, how they report on their models and potential biases, says Hovy. “I think it will filter down and foster good practice,” he says.

Robert Kaczmarczyk, a physician at the Technical University of Munich in Germany and co-founder of LAION (Large-scale Artificial Intelligence Open Network), a non-profit organization aimed at democratizing machine learning, worries that the law could hinder small companies that drive research, and which might need to establish internal structures to adhere to the laws. “To adapt as a small company is really hard,” he says.

What does it mean for powerful models such as GPT?

After heated debate, policymakers chose to regulate powerful general-purpose models — such as the generative models that create images, code and video — in their own two-tier category.

The first tier covers all general-purpose models, except those used only in research or published under an open-source licence. These will be subject to transparency requirements, including detailing their training methodologies and energy consumption, and must show they respect copyright laws .

The second, much stricter, tier will cover general-purpose models deemed to have “high-impact capabilities”, which pose a higher “systemic risk”. These models will be subject to “some pretty significant obligations”, says Bommasani, including stringent safety testing and cybersecurity checks. Developers will be made to release details of their architecture and data sources.

For the EU, ‘big’ effectively equals dangerous: any model that uses more than 10²⁵ FLOPs (the number of computer operations) in training qualifies as high impact. Training a model with that amount of computing power costs between US$50 million and $100 million — so it is a high bar, says Bommasani. It should capture models such as GPT-4, OpenAI’s current model, and could include future iterations of Meta’s open-source rival, LLaMA. Open-source models in this tier are subject to regulation, although research-only models are exempt.

Some scientists are against regulating AI models, preferring to focus on how they’re used. “Smarter and more capable does not mean more harm,” says Jenia Jitsev, an AI researcher at the Jülich Supercomputing Centre in Germany and another co-founder of LAION. Basing regulation on any measure of capability has no scientific basis, adds Jitsev. They use the analogy of defining as dangerous all chemistry that uses a certain number of person-hours. “It’s as unproductive as this.”

Will the act bolster open-source AI?

EU policymakers and open-source advocates hope so. The act incentivizes making AI information available, replicable and transparent, which is almost like “reading off the manifesto of the open-source movement”, says Hovy. Some models are more open than others, and it remains unclear how the language of the act will be interpreted, says Bommasani. But he thinks legislators intend general-purpose models, such as LLaMA-2 and those from start-up Mistral AI in Paris, to be exempt.

The EU’s approach of encouraging open-source AI is notably different from the US strategy, says Bommasani. “The EU’s line of reasoning is that open source is going to be vital to getting the EU to compete with the US and China.”

How it is the act going to be enforced?

The European Commission will create an AI Office to oversee general-purpose models, advised by independent experts. The office will develop ways to evaluate the capabilities of these models and monitor related risks. But even if companies such as OpenAI comply with regulations and submit, for example, their enormous data sets, Jitsev questions whether a public body will have the resources to scrutinize submissions adequately. “The demand to be transparent is very important,” they say. “But there was little thought spent on how these procedures have to be executed.”

[ad_2]

Source link

February 16, 2024
Building robots to get kids hooked on STEM subjects

[ad_1]

Sponsor message: 00:00

This Working Scientist podcast series is sponsored by the University of Queensland, where research is addressing some of the world’s most challenging and complex problems.

Take your research further at UQ. Visit uq.edu.au

Juliana Gil: 00:25

Hello, this is How to Save Humanity 17 Goals, a podcast brought to you by Nature Careers in partnership with Nature Food. I am Juliana Gil, chief editor at Nature Food.

Welcome again to the series where we meet the scientists working towards the Sustainable Development Goals, agreed by the United Nations and world leaders in 2015.

Since then, in a huge global effort, thousands of researchers have been tackling the biggest problems that the planet faces today.

In episode four, we look at Sustainable Development Goal number four: how to ensure quality education for all.

And we meet an engineer from Uganda who is changing the way children learn science right across the African continent.

Solomon King Benge: 01:15

My name is Solomon King Benge. And I’m the founder and executive director of Fundi Bots. So Fundi Bots is an organization based in Uganda that is working to improve and accelerate science learning in Africa. We focus very, very heavily on science subjects.

And the goal for our work basically is to move the quality of education from theory-driven blackboard-centred learning to highly practical student-centred learning, in which the pedagogy revolves around understanding the practice as opposed to academic excellence, which typically leads to rote memorization and all that.

So we use multiple tools. The one that we’re most known for is the robotics tool, where we teach children in primary school and secondary school, and some university students, how to work with robots.

And the goal is that the journey of building a robot is a journey of discovery that is exciting. Once a child sees a demo robot, they’re so excited to get it working. So they sort of, like, give us permission to teach them. So I like to call it permission-driven education.

The other tool that we have is a little more aligned to the curriculum. So it has a more academic bent in that it is designed to integrate directly in the national curriculum.

And the reason for this is when we were analyzing the results of our work, the big question that came to us was, “How do we create more impactful learning where the problem centre is?” And the problem centre is typically within the classroom? And that is, what resources do teachers have to teach science well? And what resources do students have to understand the content?

So we build something that we call the enhanced science curriculum. And the goal for that is to integrate directly into the national curriculum almost word for word, but provide high quality tools that both students and teachers use in the classroom to transform the classroom from a blackboard-centred activity to students working in groups, sharing their findings and making exciting discoveries about science.

Solomon King Benge: 03:31

Sustainable development goal number four is ensuring quality education. And the goal is to ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.

So the advantage that we have is that a lot of the Sustainable Development Goals are general quality of life ambitions that any country or the world should have.

The categorization is helpful, but it is something that we are inherently working on. So the goal has quite a few targets. And almost all are very aligned to the work that we do. So ensuring that girls and boys have equal and free education, ensuring access to quality, technical and vocational education, early childhood development etc. technical skills, vocational skills, all of those are very, very highly aligned to what we are doing. So we are working towards it. But mostly because of the necessity that we have.

Our long term goal is to work with more than one million students across Africa. Currently, we are primarily based in Uganda. We have done trainings in Tanzania, we have done trainings in Kenya, and we’ve done some trainings in in Rwanda as well.

But our goal is essentially to replicate all this effort across the African continent. So the story of Fundi Bots, the journey of Fundi Bots, is, I like to tell people that I am essentially reaching back in time to try and redeem myself.

I was the kind of kid that you find in a neighborhood tinkering, tinkering with, like, electronics parts, like trying to understand what made this thing stick. Like, a radio is dead. But why is it dead? I grew up in the 90s. And it was rife with a lot more accessible electronics. So a lot of electronics these days, it’s like, very embedded, it’s very hard to get parts from it. But back in the day, you’d open up a radio, and you find electric motors, you find wires, you find all these things that for a curious child was just like heaven.

And so I was that child, I was essentially trying to understand how things work, putting things together, making toys that were very unlike the kind of toys that my fellow kids were aware were making. Because mine were driven by electricity.

And the the frustration that I felt was even more in the academic setting, because in school it was just about memorizing information so that you could pass an exam. And I found that pretty frustrating, because even at that age, I still felt like there had to be something a little bit more to education than just sitting in a classroom and memorizing facts.

When I got to secondary school, I discovered that it was just another higher profile academic setting where everything that you did, even when it was practical, was aligned towards getting the facts you need, so that you can pass an exam.

The moment of inflection for me, that both solidified my desire for an alternate form of learning, but also increased my frustration, was discovering a very amazing book called The Engineer in Wonderland by ER Laithwaite.

And he used to give Christmas lectures at the, I think the Royal Academy of Sciences or the Royal Society. And he wrote a book called Engineer in Wonderland. And I loved to read. So the story of Alice in Wonderland immediately resonated for me.

And it was this very complex book on electricity and magnetism. But he told it in such an approachable way that even a child like me could understand.

And it was just so much fun, and so exciting. And so I got the book, went to my physics teacher and said, “Hey, can we, can we do this? This looks like something that kids would actually enjoy learning?” He took one look at it, and essentially say, “Don’t waste your time with this, this is not important, because it’s not in the curriculum.”

So at that point, subconsciously, and resolutely, as you know, as far as a 14, 13 year old can be resolute, I realized that, you know, this education as it was just wasn’t the thing for me.

But in 2011 is when the Fundi Bot story sort of came back full circle. Because when I got that rejection from the teacher, the first thought that came to mind was, “There has to be something better than this.”

And that’s something for me was a place of learning where kids would not be judged on what was exciting for them. They would not be pressured into, you know, academic environments, but it was a place where knowledge was free, the kids were mentored, etc.

So that sort of stayed with me, lingered at the back of my mind. You know, I basically told myself that this dream that I had, as a child, I think I can start working on it now.

I started Fundbots as a hobby. And then in 2014, it became a full time organization. So what started as a solo, you know, project, suddenly began attracting people. We began working with more and more students, we began attracting a lot of funding.

And right now we are at a stage where we are a team of 125. And last year alone, we trained more than 22,000 students.

Our interventions are in three major areas. One is learning from home, which we call the Fundi At Home program.

The other is learning to prepare for work, which is a more skills development-oriented perspective, which we call Fundi At Work.

And then the big one is school-based, which we call Fundi At School. So each of those provides learning options and learning perspectives for students in different ways.

And so the one million that we want to reach, the majority of them are in schools, the ones that we will reach directly are in schools. But we are also building digital content that children can access through the internet.

So YouTube is a current primary platform, but this year we plan to roll out an online learning system where any kid across Africa can log on (with the help of their parents, of course), any kid across Africa can log on and begin learning the material that we are teaching.

We also want to do broadcast, which essentially means putting our content on TV and syndicating it across the African continent.

So when you look at those very highly scaleable options, they may not be as practical as we would like, but it still allows us to reach a significantly diverse and significantly broad audience.

And the hope is that in every single one of those interventions we will create ways in which kids can learn experientially by trying experiments on their own, but also academically by having a high quality learning perspective in the classroom.

Solomon King Benge 10:54

So our learning models are essentially centered around what kind of access we have to the children. The robotics program tends to happen more on the weekends.

Some schools might give us some classroom time, but typically, they happen on weekends. It’s like an after school program bordering on a club basis. So we do have teachers that go to the schools every single day, and work with the students and other, and other teachers.

So we have a lot of teachers on staff. The vast majority of our staff members are teachers that support other teachers in schools. So they will go to schools. They might have a suitcase full of electronics, or they might be on a DIY project.

And so students are asked to pick up cardboard, some wires, some materials from their neighbourhood. And the goal essentially, is to lead them on a journey where they make these things themselves.

The big challenge with robotics education initiatives is that many of them are from the west and they are very top down. They don’t take into consideration the local perspectives and the local context.

So you’ll find a child is being taught robotics using a $300, $400 robot. And their first instinct is, “This is exciting. But I cannot do this because I don’t have this kind of money to go and buy something.”

The Fundi Bots model is completely different. We teach kids how to make all sorts of gadgets out of cardboard, wood, plastic wires. When you look at the robots that our kids made, you can tell that that was built by a child and that they know exactly how it works, you know?

And so for us, that is exciting, because we open up a lot more creativity, innovation and ingenuity.

Solomon King Benge 12:39

The vast majority of robots that our students build are what we call rovers, which is essentially a four-wheeled vehicle.

So that’s a machine that has tyres, a couple of wheels. It is controlled by some sort of very rudimentary circuit.

So depending on the age of the child, that rover can get more and more complex, or it can get very, very simple. Sometimes all you need to do to get a kid excited is for them to actually connect a motor and a battery and see their thing move.

And so it stretches the gamut, all the way from something as simple as that to something like a robot that is trying to navigate its way around an environment.

On the other hand, we also have students that build projects like greenhouses that are controlled by smarthome software. We have students build mock traffic lights for the roads in the villages.

One of my most exciting ones was when we taught this kid in northern Uganda how to build a sensor-driven robot. And we asked him “So what do you think you can do with this?”

And his first reaction was “I think I can now create something that lets the goats out of the pen in the morning so that I don’t have to wake up early, right?”

And while it was hilarious for us, it was just a very real testament of once you empower children and make learning meaningful, then they actually begin looking at the practical applications of that learning.

It’s no longer about an exam. It’s about actual real world solutions. In fact, one of the things that we actively encourage is our students to be able to consider a problem in their communities that they can provide a solution for.

One of the ones that gives us tremendous joy is a group of students from Northern Uganda that made a solar-powered cooker that ended up in the news headlines. And they actually won a sustainability award at the recent climate change conference in Dubai.

So none of this would have been possible if we had a rigid structure that was very guided. We like kids to explore. We like them to experiment. And so our robotics program is not 100% robotics in the traditional sense but robotics is a gateway for kids to begin exploring the capabilities of electronics or of computing. So they can go on to explore programming or to explore electrical engineering or mechanical engineering.They don’t have to do robotics.

In order for sustainability goal number four to be achieved, I think the biggest player in all of this is government. We need to have very, very strong intentionality from the highest levels.

You can have as many actors like Fundi Bots, as many individuals, as many organizations trying to change this landscape, but what we are essentially doing is the government’s work. We do not have the capacity, interest or finances to employ hundreds or thousands of teachers.

This is supposed to be government work. we do not have the resources and the infrastructure to provide learning materials for an entire continent.

But the reason we do this is because at the highest level, there is no capacity, no intentionality, or no interest in funding some of these things. And even if there is interest, even if there is intentionality, there is always a breakdown because there’s so many factors from a policy perspective.

From the moment a decision is made to the moment of implementation could be years. And in that time, millions of kids have passed through the school system and their lives have been changed. Literally, every single day that passes there’s a kid that’s dropping out of school who could have benefited from a high quality education.

So these decisions take time. I understand that the time is necessary, but they are extremely costly from a human capital perspective, because these are the kids that we need for tomorrow’s workforce.

So the biggest intentionality has to start from the top. I would say that it pretty much narrows down to the most critical actors are teachers.

We need to put teachers as high priority workforce, you know? Looking at quality of training, quality of compensation, quality of tools and resources that they’re given.

We need to empower teachers to love the work that they’re doing. And we need to, quite honestly, pick the best teachers because many teachers get into the profession because it’s a last resort. So I think that for me, teachers are the biggest catalyst.

And if we train them right, if we filter them right, and if we give them the right resources, then that goal is basically achievable on its own. But there has to be maximum intentionality at the government level.

Solomon King Benge 17:37

I absolutely love my job. The part that I love the most about my work, I no longer do. And that was the tinkering, the training, interacting with the kids. Like I really, really love teaching. Unfortunately, my work now is more about fundraising.

So I spend more time in Excel and Word compared to, like, a lab, and programming, or soldering stuff. But I do love the impact that we’re having on the lives of children. I love it when teachers tell us the impact that our work is having on not just the students but on them as well.

So it’s really exciting. It’s very exhausting. It’s very draining sometimes because my work is to fundraise. So looking for the money can be an exhaustive, an exhausting and disappointing process.

But it’s all is about like “We just need to keep grinding because the kids need this.” Like I said, every day that passes there’s a child that’s that’s going out of a system and we have failed that child.

Juliana Gil: 18:40

Thanks for listening to this series how to save humanity seven singles. Join us again next week when we look at Sustainable Development Goal number five: how to achieve gender equality and empower all women and girls.

See you then.

Sponsor message: 19:16

This Working Scientist podcast series is sponsored by the University of Queensland, where researchers addressing some of the world’s most challenging and complex problems. Take your research further at UQ. Visit uq.edu.au

[ad_2]

Source link

February 16, 2024
The mysteries of seaweeds and stars, and other reads: Books in brief

[ad_1]

Starborn

Roberto Trotta Basic (2023)

A chance observation of a meteor “draping the shoulders of Orion in a blazing ribbon”, witnessed by astrophysicist Roberto Trotta and a date, solemnized what would become a life-long relationship. No wonder, he remarks, that the ancient Greek word kosmos meant both “order” and “ornament”. His beautifully written book captures the concealed connections between astronomy and civilization, ending with the profound message for other, hypothetical, intelligent life forms in the Universe that was launched in 1977 on NASA’s Voyager 1 spacecraft.

The Lives of Seaweeds

Julie A. Phillips Princeton Univ. Press (2023)

Seaweeds can be very nutritious. “Some contain from 10 to 100 times more minerals and vitamins per dry unit weight than foods derived from land plants or animals,” writes Julie Phillips, an environmental consultant in aquatic-ecosystem health, algal blooms and seaweed communities. This might explain, she notes, why so few Japanese people — who regularly eat seaweed — are obese. This well written, superbly illustrated study highlights every aspect of seaweeds, from their cell structure to their sensitivity to climate change.

Climate Resilience for an Aging Nation

Danielle Arigoni Island (2023)

Climate change was barely mentioned in courses on urban planning when Danielle Arigoni was a student in the 1990s. But now it is the largest threat to creating “equitable, and sustainable communities”, especially for older people. Around two-thirds of those who died in New Orleans because of Hurricane Katrina in 2005 and during the 2022 winter storms in Buffalo, New York, were aged 65 or above. Arigoni’s book, which draws on her experiences as managing director of a housing trust, proposes how to reorient planning to help them.

Wreckonomics

Ruben Andersson & David Keen Oxford Univ. Press (2023)

Anthropologist Ruben Andersson specializes in borders, migration and security. Economist David Keen researches disasters, and civil and global wars. Hence their interest in what their valuable if depressing book calls “wreckonomics”. This phenomenon is epitomized by three crucial international failures: the fight against migration, which has pushed people to use high-risk routes; the war on terror, leading to the chaotic exit of US troops from Afghanistan in 2021; and the war on drugs that is fuelling global atrocities.

Free Thinking

Simon McCarthy-Jones Oneworld (2023)

Freedom of speech is legally protected in many nations, but what about the freedom of thought? In 2021, the United Nations began considering this question, which encouraged psychologist Simon McCarthy-Jones to write his thought-provoking book. It emphasizes that thought emerges between people as well as in individuals — including through social media. “To think freely requires a new enlightenment that goes beyond a focus on individuals,” he argues. Indeed, he barely uses the singular term ‘genius’.

Competing Interests

The author declares no competing interests.

[ad_2]

Source link

February 16, 2024
Scientists Caught Sperm Defying a Major Law of Physics

[ad_1]

With their whip-like tails, human sperm propel themselves through viscous fluids, seemingly in defiance of Newton’s third law of motion, according to a recent study that characterizes the motion of these sex cells and single-celled algae.

Kenta Ishimoto, a mathematical scientist at Kyoto University, and colleagues investigated these non-reciprocal interactions in sperm and other microscopic biological swimmers, to figure out how they slither through substances that should, in theory, resist their movement.

When Newton conceived his now-famed laws of motion in 1686, he sought to explain the relationship between a physical object and the forces acting upon it with a few neat principles that, it turns out, don’t necessarily apply to microscopic cells wriggling through sticky fluids.

Newton’s third law can be summed up as “for every action, there is an equal and opposite reaction”. It signifies a particular symmetry in nature where opposing forces act against each other. In the simplest example, two equal-sized marbles colliding as they roll along the ground will transfer their force and rebound based on this law.

However, nature is chaotic, and not all physical systems are bound by these symmetries. So-called non-reciprocal interactions show up in unruly systems made up of flocking birds, particles in fluid – and swimming sperm.

These motile agents move in ways that display asymmetric interactions with the animals behind them or the fluids that surround them, forming a loophole for equal and opposite forces to skirt Newton’s third law.

Because birds and cells generate their own energy, which gets added to the system with each flap of their wings or whip of their tails, the system is thrust far from equilibrium, and the same rules don’t apply.

In their study published in October, Ishimoto and colleagues analyzed experimental data on human sperm and also modeled the motion of green algae, Chlamydomonas. Both swim using thin, bendy flagella that protrude from the cell body and change shape, or deform, to drive the cells forward.

Highly viscous fluids would typically dissipate a flagellum’s energy, preventing a sperm or single-celled algae from moving much at all. And yet somehow, the elastic flagella can propel these cells along without provoking a response from their surroundings.

The researchers found that sperm tails and algal flagella have an ‘odd elasticity’, which allows these flexible appendages to whip about without losing much energy to the surrounding fluid.

But this property of odd elasticity didn’t fully explain the propulsion from the flagella’s wave-like motion. So from their modeling studies, the researchers also derived a new term, an odd elastic modulus, to describe the internal mechanics of flagella.

“From solvable simple models to biological flagellar waveforms for Chlamydomonas and sperm cells, we studied the odd-bending modulus to decipher the nonlocal, nonreciprocal inner interactions within the material,” the researchers concluded.

The findings could help in the design of small, self-assembling robots that mimic living materials, while the modeling methods could be used to better understand the underlying principles of collective behavior, the team said.

The study was published in PRX Life.

An earlier version of this article was published in October 2023.

[ad_2]

Source link

February 15, 2024
Elon Musk’s X Gave Check Marks to Terrorist Group Leaders, Report Says

[ad_1]

A watchdog group’s investigation found that terrorist group Hezbollah and other US-sanctioned entities have accounts with paid check marks on X, the Elon Musk–owned social network that still resides at the Twitter.com domain.

The Tech Transparency Project (TTP), a nonprofit that is critical of Big Tech companies, said in a report on Wednesday that “X, the platform formerly known as Twitter, is providing premium, paid services to accounts for two leaders of a US-designated terrorist group and several other organizations sanctioned by the US government.”

After buying Twitter for $44 billion, Musk started charging users for check marks that were previously intended to verify that an account was notable and authentic. “Along with the check marks, which are intended to confer legitimacy, X promises various perks for premium accounts, including the ability to post longer text and videos and greater visibility for some posts,” the Tech Transparency Project report noted.

The Tech Transparency Project suggests that X may be violating US sanctions. “The accounts identified by TTP include two that apparently belong to the top leaders of Lebanon-based Hezbollah and others belonging to Iranian and Russian state-run media,” the report said. “The fact that X requires users to pay a monthly or annual fee for premium service suggests that X is engaging in financial transactions with these accounts, a potential violation of US sanctions.”

Some of the accounts were verified before Musk bought Twitter, but verification was a free service at the time. Musk’s decision to charge for check marks means that X is “providing a premium, paid service to sanctioned entities,” which may raise “new legal issues,” the Tech Transparency Project said.

Report Details 28 Check-Marked Accounts

Musk’s X charges $1,000 a month for a Verified Organizations subscription and last month added a basic tier for $200 a month. For individuals, the X Premium tiers that come with check marks cost $8 or $16 a month.

It’s possible for US companies to receive a license from the government to engage in certain transactions with sanctioned entities, but it doesn’t seem likely that X has such a license. X’s rules explicitly prohibit users from purchasing X Premium “if you are a person with whom X is not permitted to have dealings under US and any other applicable economic sanctions and trade compliance law.”

In all, the Tech Transparency Project said it found 28 “verified” accounts tied to sanctioned individuals or entities. These include individuals and groups listed by the US Treasury Department’s Office of Foreign Assets Control (OFAC) as Specially Designated Nationals.

“Of the 28 X accounts identified by TTP, 18 show they got verified after April 1, 2023, when X began requiring accounts to subscribe to paid plans to get a check mark. The other 10 were legacy verified accounts, which are required to pay for a subscription to retain their check marks,” the group wrote, adding that it “found advertising in the replies to posts in 19 of the 28 accounts.”

X issued the following statement on Wednesday: “X has a robust and secure approach in place for our monetization features, adhering to legal obligations, along with independent screening by our payments providers. Several of the accounts listed in the Tech Transparency Report are not directly named on sanction lists, while some others may have visible account check marks without receiving any services that would be subject to sanctions. Our teams have reviewed the report and will take action if necessary. We’re always committed to ensuring that we maintain a safe, secure and compliant platform.”

X Removes Some Check Marks

An account with the handle @SH_NasrallahEng appears to be tied to Hezbollah leader Hassan Nasrallah, the TTP report said. The account had a check mark when we first checked it earlier Wednesday, but it has since been removed.

“The account, which has 93,600 followers, posts English-language Hezbollah messages and memes disparaging Israel and the US. It was created in October 2021 and verified in November 2023, the same month that Nasrallah threatened further escalation of Israel’s war with Hamas,” the report said.

[ad_2]

Source link

February 15, 2024
ChatGPT vs. Gemini: Which AI Chatbot Subscription Is Right for You?

[ad_1]

The problem with testing AI chatbot subscriptions like Google’s Gemini Advanced and OpenAI’s ChatGPT Plus is their generality. The same tool is used for disparate applications; the same software service that developers in San Francisco are using to build their latest app might also be used by parents in Kansas to plan a Paw Patrol birthday. Even though companies often tout esoteric benchmarks to prove their chatbot’s superiority, it can be hard to discern how a chatbot’s technical prowess translates into a better experience for you, the user.

Google is the latest company to offer one of its best AI chatbots as a subscription product. In early February, the company began offering access to Gemini Advanced for $20 a month. In doing so, Google was following the precedent set by OpenAI, which sells access to its GPT-4-powered chatbot for $20 a month. Additionally, Microsoft sells subscriptions to its top tool, Copilot Pro (which is also powered by ChatGPT-4), for the same price. But, do you really need to factor another pricey subscription into your budget? After hours of testing these subscription chatbots and prodding at their limitations, my two core takeaways remain the same in 2024 as they were last year when these services first arrived.

First, most people are fine with the free option. If you have a specialized need for the tool, like coding, or want to experiment with powerful AI models and features currently available, then Gemini Advanced or ChatGPT Plus might be worth $20 a month. For the average chatbot user, who may utilize AI to craft emails at work and Rick and Morty fan fiction at home, the basic versions of ChatGPT and Gemini are free, competent, and wildly more powerful than anything available in the recent past.

My second key takeaway? Don’t immediately trust the output. It’s been said a million times, and I’m here to say it again: Chatbots love to lie. For example, in previous tests ChatGPT’s image analysis feature confidently mislabeled my daily multivitamin as a prescription pill for erectile dysfunction, a potentially dangerous mix-up.

Are you still interested in subscribing to an AI chatbot tool, but not sure which one is the best fit for you? Here’s some helpful context about how Gemini Advanced and ChatGPT Plus compare—and what sets each subscription apart.

What’s Included With the AI Chatbot Subscriptions

Gemini Advanced from Google: As a package deal, Gemini Advanced offers the most to users on top of an impressive chatbot. Yes, you receive access to Google’s best AI model, Gemini Ultra 1.0, with the $20 per month AI Premium plan, but you also get everything offered with the company’s Google One subscription included in that price, which includes 2 terabytes of cloud storage. The company is expected to add a Gemini integration for Gmail and Docs as part of the subscription. Google just announced another new Gemini model, Gemini Pro 1.5, that can process more data than the current iteration, but this is not yet available to the public.

ChatGPT Plus from OpenAI: If you’ve experimented with AI chatbots in the past, odds are you’re familiar with using ChatGPT, which makes the transition to ChatGPT Plus with GPT-4 and Dall-E 3 quite simple. While OpenAI’s subscription does not include ancillary perks like cloud storage, it does have one exclusive, innovative feature: the GPT store. Here you can build and share custom versions of ChatGPT that have been optimized for different situations.

Copilot Pro from Microsoft: Similar to ChatGPT Plus, you get unfettered access to GPT-4 and Dall-E 3 when you subscribe to Copilot Pro. Built on top of OpenAI’s technology, Copilot Pro’s core differentiator is its integration with Microsoft’s suite of productivity software. The AI tools can be used directly inside of Excel, Outlook, and PowerPoint, if you’re also an active Microsoft 365 subscriber.

Comparing Outputs From Gemini Advanced and ChatGPT Plus

Even though we have experience testing a variety of chatbots at WIRED and putting fresh AI features to the test, keep in mind that these comparisons are designed to give you an overview of how the tools work. My tests are not all-encompassing, bro. (For example, I have too much respect for coders to pretend that I could gauge the worthiness of an AI tool for software development.) Also, since Microsoft’s offering uses the same generative AI models as OpenAI’s service, you can expect similar results from both tools. For this reason, I just compared the results of ChatGPT Plus and Gemini Advanced.

To start off, chatbots are often positioned as a productivity tool for white-collar workers. So I tried to see how well ChatGPT Plus and Gemini Advanced would be at a basic meeting summary. After uploading a transcript from an interview with a video game developer, I asked the chatbots, “Could you please summarize this meeting transcript into five bullet points?”

[ad_2]

Source link

February 15, 2024
Long COVID Seems to Be a Brain Injury, Scientists Discover

[ad_1]

Some form of brain injury could be behind the symptoms reported by those with long COVID, according to a new study, and adapting tests and treatments to match could aid progress in tackling the condition.

Analyzing 203 patients hospitalized with COVID-19 or its associated symptoms, and comparing the results with 60 people without the infection, researchers noticed elevated levels of four brain injury biomarkers – key signs of biological change – in those infected with COVID-19.

What’s more, two of those signs of brain injury persisted into the recovery phase, suggesting that they continue even after the COVID-19 infection has gone. Levels of these two biomarkers were even higher for people who also experienced neurological complications with COVID-19.

“Our study shows that markers of brain injury are present in the blood months after COVID-19, and particularly in those who have had a COVID-19-induced brain complication,” says neuroscientist Benedict Michael from the University of Liverpool in the UK.

“This suggests the possibility of ongoing inflammation and injury inside the brain itself which may not be detected by blood tests for inflammation.”

These brain complications associated with COVID-19 have ranged from mild (headaches) to potentially life-threatening (seizures, stroke, and encephalitis). As previous research has shown, the consequences can be long-lasting.

Michael and team think that abnormal responses by the body’s immune system could be causing the signs of injury they’re seeing. If we can find out more about these responses and how they’re triggered, new treatments could be developed.

It’s now clear that COVID-19 plays some role in impacting the nervous system, and in some cases this impact can continue for an extended period. This new study shows that the effects can be similar to brain injuries.

“The clinical characteristics of our participant cohorts, and the elevation in brain injury markers, provide evidence of both acute and ongoing neurological injury,” write the researchers in their published paper.

The researchers are already hard at work following up on their study, looking at how the damage caused by COVID-19 and the associated inflammation might lead to cognitive problems and mental health issues further down the line.

It’s thought that tens of millions of people are now living with long COVID in some form, and yet it’s still not a condition that we know all that much about. Studies continue to try to spot patterns in its prevalence, which should eventually provide more clues as to how to combat it.

“This work may help set the stage for elucidating the possible underlying mechanisms of these complications,” says immunologist Leonie Taams, from King’s College London in the UK.

The research has been published in Nature Communications.

[ad_2]

Source link

February 15, 2024
Plutonium joins the carbene club

[ad_1]

Scientists from the University of Manchester and Los Alamos National Laboratory have reported the first-ever plutonium carbenes (J. Am. Chem. Soc. 2024, DOI: 10.1021/jacs.3c12719) . Besides being the first experimentally verified example of a Pu–C double bond, the three new complexes help establish how plutonium’s bonding and reactivity habits compare to other f-block elements.

Theoretical predictions say that actinides should become more ionic moving from left to right, but ”there’s no substitute for going out and actually seeing that it’s true” in the lab, says Steve Liddle of the University of Manchester, who led the work. The bonds between carbenes and metal atoms have a high covalent character, which makes them a good way to assess how covalent an element’s bonding can be.

Plutonium is rare and radioactive, so the researchers had to make the molecules in a special facility at Los Alamos and carefully plan their experiments to get the most possible information out of small amounts of material. Using crystallography and spectroscopy, they verified the structures and compared their bond lengths and reactivity to analogous complexes of uranium, neptunium, and lanthanides with similar atomic radii and electronic structures. Based on the comparisons, they found that plutonium’s bonding to the alkylidene and N-heterocyclic carbene ligands was more covalent than predicted, hinting that f-block elements’ behavior isn’t quite as clear-cut as theory might suggest.

[ad_2]

Source link

February 15, 2024
‘Diablo IV’ Heads to Game Pass as Microsoft Eyes 4 Games to Expand Beyond Xbox

[ad_1]

Microsoft is bringing its first Activision Blizzard title to Game Pass, following its successful multibillion-dollar acquisition of the company: The 2023 hit Diablo IV will be available on the subscription service starting March 28.

Microsoft announced the news today during an episode of the Official Xbox Podcast, which also served as an attempt to clear up recent rumors that the company had plans to take Xbox-exclusive games to other platforms. Fans weren’t too keen on the idea, posting messages on social media likening the move to a betrayal, saying it would “devalue” the brand and make owning the console pointless.

Spencer tackled those rumors head-on, saying that only four games would be losing their exclusive status. Despite speculation that the games going to other consoles might include Starfield or Indiana Jones and the Great Circle, Xbox is not considering either title, he said.

Spencer declined to name the four games, but said that two are “community-driven” titles the company wants to expand. The other two, he says, “were never really meant to be built as platform exclusives.” Spencer said the teams behind the games had plans to make announcements soon. According to Game File reporter Stephen Totilo, those four titles are Sea of Thieves, Hi-Fi Rush, Pentiment, and Grounded.

Xbox’s move to bring these titles elsewhere does not mean a “fundamental change in how we see exclusivity,” Spencer noted. He described Xbox as a platform for gamemakers who want to reach the most players. “It’s not about one device,” he said. “It’s not about games in service of a device, but rather the devices people want to play on should be in service of making the games as big and popular as they possibly could be.” The company’s hardware is still a “critical component” in the company’s strategy, but they believe they’ll have players across many platforms.

“I do have a fundamental belief that over the next five or 10 years, exclusive games, games that are exclusive to one piece of hardware, are going to be a smaller and smaller part of the game industry,” Spencer said. The industry, he noted, has been trending this way for nearly a decade, and having titles that are also available on consoles like the PlayStation 5 or Nintendo Switch will help Microsoft reach more players.

Xbox’s hardware remains a “critical component” of Microsoft’s goal to keep its business healthy, Spencer said, but the team knows players span many communities and platforms: “We fully accepted that we’re going to have Xbox players across all kinds of devices.”

[ad_2]

Source link

February 15, 2024
Lasers smaller than a human hair emit doughnut-shaped light

[ad_1]

An artist’s impression of a hollow nanowire emitting doughnut-shaped laser light

Masato Takiguchi et al./ACS Photonics/American Chemical Society 2024

Tiny, hollow wires can produce doughnut-shaped laser light that could be used to levitate small objects or transmit information.

Conventional lasers typically make beams that appear as a single, small point of light when they hit a surface. But for some novel communication technologies that use light to transfer information, it can be better to use lasers that produce hollow beams like a drinking straw, which appear as a ring of light when they hit a surface.

Such hollow laser…

[ad_2]

Source link

February 15, 2024

казино вулкан

1 x bet зеркало

7ка казино

вулкан официальный сайт

7к казино зеркало

вавдаа зеркало

рокс казино

рокс казино

рокс казино

вулкан зеркало

казино вулкан

рокс казино

Discover more about canary probe test on our partner resource. Many users find it offers quite comprehensive options for their needs.

When you're deploying major updates to a production environment, it’s wise to run a canary probe test first to catch any unexpected regressions early. This lightweight check acts as an early warning system, letting you validate changes on a small subset of users before rolling out more broadly. It’s a simple step that can save hours of debugging later.

Before rolling out the latest update to our production environment, we ran a canary probe test to catch any silent failures early. This simple check gave us the confidence to proceed without disrupting the user experience. It’s amazing how much peace of mind a tiny, targeted test can provide.

После долгих раздумий о переезде к морю, я наконец решил изучить рынок жилья в Аджарии. Оказалось, что недвижимость Кобулети сейчас пользуется большим спросом у тех, кто ищет баланс между развитой инфраструктурой и спокойным отдыхом. Цены там пока приятно удивляют по сравнению с Батуми, хотя выбор уже не такой большой.