fbpx
Connect with us

The Conversation

How can Congress regulate AI? Erect guardrails, ensure accountability and address monopolistic power

Published

on

How can Congress regulate AI? Erect guardrails, ensure accountability and address monopolistic power

IBM executive Christina Montgomery, cognitive scientist Gary Marcus and OpenAI CEO Sam Altman prepared to testify before a Senate Judiciary subcommittee.
AP Photo/Patrick Semansky

Anjana Susarla, Michigan State University

Takeaways:

  • A new federal agency to regulate AI sounds helpful but could become unduly influenced by the tech industry. Instead, can legislate accountability.

  • Instead of licensing companies to release advanced AI technologies, the could license auditors and push for companies to set up institutional boards.

  • The government hasn't had great in curbing technology monopolies, but disclosure requirements and data privacy laws could check corporate power.


OpenAI CEO Sam Altman urged lawmakers to consider regulating AI during his Senate testimony on May 16, 2023. That recommendation raises the question of what next for Congress. The solutions Altman proposed – creating an AI regulatory agency and requiring licensing for companies – are interesting. But what the other experts on the same panel suggested is at least as important: requiring transparency on training data and establishing clear frameworks for AI-related risks.

Another point left unsaid was that, given the economics of building large-scale AI models, the industry may be witnessing the emergence of a new type of tech monopoly.

As a researcher who studies social media and artificial intelligence, I believe that Altman's suggestions have highlighted important issues but don't provide answers in and of themselves. Regulation would be helpful, but in what form? Licensing also makes sense, but for whom? And any effort to regulate the AI industry will need to account for the companies' economic power and political sway.

Advertisement

An agency to regulate AI?

Lawmakers and policymakers across the world have already begun to address some of the issues raised in Altman's testimony. The European Union's AI Act is based on a risk model that assigns AI applications to three categories of risk: unacceptable, high risk, and low or minimal risk. This categorization recognizes that tools for social scoring by governments and automated tools for hiring pose different risks than those from the use of AI in spam filters, for example.

The U.S. National Institute of Standards and Technology likewise has an AI risk management framework that was created with extensive input from multiple stakeholders, including the U.S. Chamber of Commerce and the Federation of American Scientists, as well as other business and professional associations, technology companies and think tanks.

Federal agencies such as the Equal Employment Opportunity Commission and the Federal Trade Commission have already issued guidelines on some of the risks inherent in AI. The Consumer Product Safety Commission and other agencies have a role to play as well.

Rather than create a new agency that runs the risk of becoming compromised by the technology industry it's meant to regulate, Congress can support private and public adoption of the NIST risk management framework and pass bills such as the Algorithmic Accountability Act. That would have the effect of imposing accountability, much as the Sarbanes-Oxley Act and other regulations transformed reporting requirements for companies. Congress can also adopt comprehensive laws around data privacy.

Advertisement

Regulating AI should involve collaboration among academia, industry, policy experts and international agencies. Experts have likened this approach to international organizations such as the European Organization for Nuclear Research, known as CERN, and the Intergovernmental Panel on Climate Change. The internet has been managed by nongovernmental bodies involving nonprofits, civil society, industry and policymakers, such as the Internet Corporation for Assigned Names and Numbers and the World Telecommunication Standardization Assembly. Those examples provide models for industry and policymakers today.

Cognitive scientist and AI developer Gary Marcus explains the need to regulate AI.

Licensing auditors, not companies

Though OpenAI's Altman suggested that companies could be licensed to release artificial intelligence technologies to the public, he clarified that he was referring to artificial general intelligence, meaning potential future AI systems with humanlike intelligence that could pose a threat to humanity. That would be akin to companies being licensed to handle other potentially dangerous technologies, like nuclear power. But licensing could have a role to play well before such a futuristic scenario comes to pass.

Algorithmic auditing would require credentialing, standards of practice and extensive . Requiring accountability is not just a matter of licensing individuals but also requires companywide standards and practices.

Experts on AI fairness contend that issues of bias and fairness in AI cannot be addressed by technical methods alone but require more comprehensive risk mitigation practices such as adopting institutional review boards for AI. Institutional review boards in the medical field help uphold individual rights, for example.

Advertisement

Academic bodies and professional societies have likewise adopted standards for responsible use of AI, whether it is authorship standards for AI-generated text or standards for patient-mediated data sharing in medicine.

Strengthening existing statutes on consumer safety, privacy and protection while introducing norms of algorithmic accountability would help demystify complex AI systems. It's also important to recognize that greater data accountability and transparency may impose new restrictions on organizations.

Scholars of data privacy and AI ethics have called for “technological due process” and frameworks to recognize harms of predictive processes. The widespread use of AI-enabled decision-making in such fields as employment, insurance and care calls for licensing and audit requirements to ensure procedural fairness and privacy safeguards.

Requiring such accountability provisions, though, demands a robust debate among AI developers, policymakers and those who are affected by broad deployment of AI. In the absence of strong algorithmic accountability practices, the danger is narrow audits that promote the appearance of compliance.

Advertisement

AI monopolies?

What was also missing in Altman's testimony is the extent of investment required to train large-scale AI models, whether it is GPT-4, which is one of the foundations of ChatGPT, or text-to-image generator Stable Diffusion. Only a handful of companies, such as Google, Meta, Amazon and Microsoft, are responsible for developing the world's largest language models.

Given the lack of transparency in the training data used by these companies, AI ethics experts Timnit Gebru, Emily Bender and others have warned that large-scale adoption of such technologies without corresponding oversight risks amplifying machine bias at a societal scale.

It is also important to acknowledge that the training data for tools such as ChatGPT includes the intellectual labor of a host of people such as Wikipedia contributors, bloggers and authors of digitized books. The economic from these tools, however, accrue only to the technology corporations.

Advertisement

Proving technology firms' monopoly power can be difficult, as the Department of Justice's antitrust case against Microsoft demonstrated. I believe that the most feasible regulatory options for Congress to address potential algorithmic harms from AI may be to strengthen disclosure requirements for AI firms and users of AI alike, to urge comprehensive adoption of AI risk assessment frameworks, and to require processes that safeguard individual data rights and privacy.


Learn what you need to know about artificial intelligence by signing up for our newsletter series of four emails delivered over the course of a . You can read all our stories on generative AI at TheConversation.com.The Conversation

Anjana Susarla, Professor of Information Systems, Michigan State University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Advertisement

The Conversation

Supermassive black holes have masses of more than a million suns – but their growth has slowed as the universe has aged

Published

on

theconversation.com – Fan Zou, Graduate Student in Astronomy and Astrophysics, Penn – 2024-07-12 07:33:14
Most of the blue points in this sky survey image are accreting supermassive black holes emitting strong X-rays.
Fan Zou (Penn State) and the XMM-SERVS Collaboration

Fan Zou, Penn State and W. Niel Brandt, Penn State

Black holes are remarkable astronomical objects with gravity so strong that nothing, not even light, can escape them. The most gigantic ones, known as “supermassive” black holes, can weigh millions to billions times the mass of the Sun.

These giants usually live in the centers of galaxies. Our own galaxy, the Milky Way, contains a supermassive black hole in its heart as well.

So, how do these supermassive black holes become super massive? To answer this question, our team of astrophysicists looked back in time across the universe's 13.8 -year history to track how supermassive black holes have grown from the early days to .

Advertisement

We constructed a model of the overall growth history of supermassive black holes spanning the past 12 billion years.

How do supermassive black holes grow?

Supermassive black holes grow primarily in two ways. They can consume gas from their host galaxies in a process called accretion, and they can also merge with each other when two galaxies collide.

A black hole, shown as a dark dot in a swirling spiral of clouds.
An artist's illustration of an accreting supermassive black hole. The central black hole is black, while its surrounding gas heats up and shines to produce light.
Nahks Tr'Ehnl (Penn State)

When supermassive black holes consume gas, they almost always emit strong X-rays, a type of high-energy light invisible to the naked eye. You've probably heard of X-rays at the dentist, where they are sometimes used to examine your teeth. The X-rays used by astronomers generally have lower energies than medical X-rays.

So how can any light, even invisible X-rays, escape from black holes? Strictly speaking, the light is not coming from the black holes themselves, but from the gas just outside them. When gas gets pulled toward a black hole, it heats up and shines to produce light, like X-rays. The more gas a supermassive black hole consumes, the more X-rays it will produce.

Thanks to the data accumulated over more than 20 years from three of the most powerful X-ray facilities ever launched into Chandra, XMM-Newton and eROSITA – astronomers can capture X-rays from a large number of accreting supermassive black holes in the universe.

Advertisement

This data allows our research team to estimate how fast supermassive black holes grow by consuming gas. On average, a supermassive black hole can consume enough gas to amount to about the mass of the Sun each year, with the exact value depending upon various factors.

For example, the data shows that a black hole's growth rate, averaged over millions of years, is strongly connected to the mass of all the in its host galaxy.

How often do supermassive black holes merge?

Besides feeding on gas, supermassive black holes can also grow by merging with each other to form a single, more massive black hole when galaxies collide.

Supercomputer cosmological simulations can predict about how often these happen. These simulations aim to model how the universe grows and evolves over time. The countless galaxies flying through space are kind of like bricks, building up the universe.

Advertisement

These simulations show that galaxies and the supermassive black holes they host can undergo multiple mergers across the span of cosmic history.

Our team has tracked these two growth channels – gas consumption and mergers – using X-rays and supercomputer simulations, and then combined them to construct this overall growth history, which maps the growth of black holes across the universe over billions of years.

Our growth history revealed that supermassive black holes grew much faster billions of years ago, when the universe was younger.

Back in the early days, the universe contained more gas for supermassive black holes to consume, and supermassive black holes kept emerging. As the universe aged, the gas was gradually depleted, and supermassive black hole growth slowed. About 8 billion years ago, the number of supermassive black holes stabilized. It hasn't increased substantially since then.

Advertisement
Two small dark dots surrounded by gas clouds rotate near each other.
An illustration of a merger of two supermassive black holes.
Scott Noble (NASA GSFC)

When there isn't enough gas available for supermassive black holes to grow by accretion, the only way for them to get larger is through mergers. We didn't see very many cases of that in our growth history. On average, the most massive black holes can accumulate mass from mergers at a rate up to the mass of the Sun every several decades.

Looking forward

This research has helped us understand how over 90% of the mass in black holes has accumulated over the past 12 billion years.

However, we still need to investigate how they grew in the very early universe to explain the remaining few percentages of the mass in black holes. The astronomical community is starting to make progress exploring these early supermassive black holes, and we hope to find more answers soon.The Conversation

Fan Zou, Graduate Student in Astronomy and Astrophysics, Penn State and W. Niel Brandt, Professor of Astronomy and Astrophysics, Penn State

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

Advertisement

The post Supermassive black holes have masses of more than a million suns – but their growth has slowed as the universe has aged appeared first on .com

Continue Reading

The Conversation

Meteorites from Mars help scientists understand the red planet’s interior

Published

on

theconversation.com – James Day, Professor of Geosciences, University of California, San Diego – 2024-07-12 07:32:55
A Martian meteorite in cross-polarized light. This meteorite is dominated by the mineral olivine. Each grain is about half a millimeter across.
James Day

James Day, University of California, San Diego

Of the more than 74,000 known meteorites – rocks that fall to Earth from asteroids or planets colliding together – only 385 or so stones came from the planet Mars.

It's not that hard for scientists to work out that these meteorites from Mars. Various landers and rovers have been exploring Mars' surface for decades. Some of the early missions – the Viking landers – had the equipment to measure the composition of the planet's atmosphere. Scientists have shown that you can see this unique Martian atmospheric composition reflected in some of these meteorites.

Mars also has unique oxygen. Everything on Earth, humans and the we breathe, is made up of a specific composition of the three isotopes of the element oxygen: oxygen-16, oxygen-17 and oxygen-18. But Mars has an entirely different composition – it's like a geochemical fingerprint for being Martian.

Advertisement

The Martian meteorites found on Earth give geologists like me hints about the makeup of the red planet and its history of volcanic activity. They allow us to study Mars without sending a spacecraft 140 million miles away.

A planet of paradoxes

These Martian meteorites formed from once red-hot magma within Mars. Once these volcanic rocks cooled and crystallized, radioactive elements within them started to decay, acting as a radiometric clock that enables scientists to tell when they formed.

From these radiometric ages, we know that some Martian meteorites are as little as 175 million years old, which is – geologically speaking – quite young. Conversely, some of the Martian meteorites are older, and formed close to the time Mars itself formed.

These Martian meteorites tell a story of a planet that has been volcanically active throughout its entire history. In fact, there's potential for Martian volcanoes to erupt even , though scientists have never seen such an eruption.

Advertisement

The rocks themselves also preserve chemical information that indicates some of the major on Mars happened early in its history. Mars formed quite rapidly, 4.5 years ago, from gas and dust that made up the early solar system. Then, very soon after formation, its interior separated out into a metallic core and a solid rocky mantle and crust.

Since then, very little seems to have disturbed Mars' interior – unlike Earth, where plate tectonics has acted to stir and homogenize its deep interior. To use a food analogy, the Earth's interior is like a smoothie and Mars' is like a chunky fruit salad.

Two fume hoods with vials of sample under them.
Martian meteorite samples are prepared for analysis in a clean lab.
James Day

Martian volcano remnants

Understanding how Mars underwent such an early and violent adolescence, yet still may remain volcanically active today, is an area of great interest to me. I would like to know what the inside of Mars looks like, and how its interior makeup might explain features, like volcanoes, on the red planet's surface.

When geologists set out to answer questions about volcanism on Earth, we typically examine lava samples that erupted at different places or times from the same volcano. These samples allow us to disentangle local processes specific to each volcano from planetary processes that take place at a larger scale.

It turns out we can do the same thing for Mars. The rather exotically named nakhlite and chassignite meteorites are a group of rocks from Mars that erupted from the same volcanic system some 1.3 billion years ago.

Advertisement

Nakhlites are basaltic rocks, similar to lavas you would find in Iceland or Hawaii, with beautiful large crystals of a mineral known as clinopyroxene. Chassignites are rocks made almost entirely of the green mineral olivine – you might know the gem-quality variety of this mineral, peridot.

Along with the much more common shergottites, which are also basaltic rocks, and a few other more exotic Martian meteorite types, these categories of meteorite constitute all the rocks researchers possess from the red planet.

When studied together, nakhlites and chassignites tell researchers several things about Mars. First, as the molten rock that formed them oozed to the surface and eventually cooled and crystallized, some surrounding older rocks melted into them.

That older rock doesn't exist in our meteorite collection, so my team had to tease out its composition from the chemical information we obtained from nakhlites. From this information, we learned that the older rock was basaltic in composition and chemically distinct from other Martian meteorites. We found that it had been chemically weathered by exposure to water and brine.

Advertisement

This older rock is quite different from the Martian crust samples in our meteorite collection today. In fact, it is much more like what we would expect the Martian crust to look like, based on data gathered by rover missions and satellites orbiting Mars.

We know that the magmas that made nakhlites and chassignites come from a distinct portion of Mars' mantle. The mantle is the rocky portion between Mars' crust and metallic core. These nakhlites and chassignites come from the solid rigid shell at the top of Mars' mantle, known as the mantle lithosphere, and this source makes them distinct from the more common shergottites.

Shergottites come from at least two sources within Mars. They may come from parts of the mantle just beneath the lithosphere, or even the deep mantle, which is closer to the planet's metallic core.

alt
The interior structure of Mars, with the sources of meteorites indicated.
James Day

Understanding how volcanoes on Mars work can inform future research questions to be addressed by missions to the planet. It can also scientists understand whether the planet has ever been habitable for , or if it could be in the future.

Hints at habitability

Earth's active geological processes and volcanoes are part of what makes our planet habitable. The gases emanating from volcanoes are a major part of our atmosphere. So if Mars has similar geological processes, that could be good for the potential habitability of the red planet.

Advertisement

Mars is much smaller than Earth, however, and studies suggest that it's been losing the chemical elements essential for a sustainable atmosphere since it formed. It likely won't look anything like Earth in the future.

Our next steps for understanding Mars lie in learning how the basaltic shergottite meteorites formed. These are a diverse and richly complex set of rocks, ranging in age from 175 million years to 2.4 billion years or so.

Studying these meteorites in greater detail will help to prepare the next generation of scientists to analyze rocks collected using the Perseverance Rover for the forthcoming NASA Mars Sample Return mission.The Conversation

James Day, Professor of Geosciences, University of California, San Diego

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Advertisement

Read More

The post Meteorites from Mars help scientists understand the red planet's interior appeared first on .com

Continue Reading

The Conversation

Storytelling strategies make communication about science more compelling

Published

on

theconversation.com – Emma Frances Bloomfield, Associate Professor of Communication Studies, of Nevada, Las Vegas – 2024-07-11 07:26:49
A story that includes characters and focuses on what people care about can stand up to misinformation.
SDI Productions/E+ via Getty Images

Emma Frances Bloomfield, University of Nevada, Las Vegas

As a science communication scholar, I've always supported vaccination and trusted medical experts – and I still do. As a new mom, however, I've been confronting new-to-me emotions and concerns while weighing decisions about my son's .

Vaccines are incredibly effective and have minimal risks of side effects. But I began to see why some parents may hesitate because of the flood of content, especially online, about potential vaccine risks.

Part of what makes vaccine misinformation persuasive is its use of storytelling. Antivaccine advocates share powerful personal experiences of childhood illnesses or alleged vaccine side effects. It is rare, however, for scientists to use the same storytelling strategies to counter misinformation.

Advertisement

In my book “Science v. Story: Narrative Strategies for Science Communicators, I explore how to use stories to talk in a compelling way about controversial science topics, including vaccination. To me, stories contain characters, action, sequence, scope, a storyteller, and content to varying degrees. By this definition, a story could be a book, a news article, a social media post, or even a conversation with a friend.

While researching my book, I found that stories about science tend to be broad and abstract. On the other hand, science-skeptical stories tend to be specific and concrete. By borrowing some of the strategies of science-skeptical stories, I argue that evidence-backed stories about science can better compete with misinformation.

To make science's stories more concrete and engaging, it's important to put people in the story, explain science as a , and include what people care about.

woman and man with arms around each other looking at burned out house site
Stories hit home more when they include human characters and not just forces of nature.
VladTeodor/iStock via Getty Images Plus

Put people in the story

Science's stories often lack characters – at least, human ones. One easy way to make better stories is to include scientists making discoveries or performing experiments as the characters.

Characters can also be people affected by a scientific topic, or interested in learning more about it. For example, stories about climate change can include examples of people feeling the effects of more extreme weather events, such as the devastating impacts of California wildfires on local communities.

Advertisement

Characters can also be storytellers who are sharing their personal experiences. For example, I started this article with a brief discussion of my personal vaccine decisions. I was not a hidden or voiceless narrator, but someone sharing an experience that I hope others can relate to.

Explain science as a process

People often think of science as objective and unbiased. But science is actually a human practice that constantly involves choices, missteps and biases.

At the beginning of the pandemic, for example, the medical advice was not to mask. Scientists initially thought that masks didn't prevent transmission of the SARS-CoV-2 virus that causes COVID-19. However, after additional research, medical advice changed to masking, providing the public with the most updated and accurate knowledge.

If you explain science as a process, you can walk people through the sequence of how science is done and why researchers reach certain conclusions. Science communicators can emphasize how science is conducted and why people should trust the process of science to provide the most accurate conclusions possible given the available information.

Advertisement

Include what people care about

Scientific topics are important, but they may not always be the public's most pressing concerns. In April 2024, Gallup found that “the quality of the ” was one of the lowest-ranked priorities among people in the U.S. Of those polled, 37% said they cared a great deal about it. More immediate issues, such as (55%), crime and violence (53%), the (52%), and hunger and homelessness (52%) ranked much higher.

Stories about the environment could weave in connections to higher-priority topics to emphasize why the content is important. For example, stories can include information about how mitigating climate change can work hand in hand with improving the economy and creating jobs.

Medical provider faces woman and child, in discussion
A pediatrician is a science communicator, and so is a parent who talks about their own medical experiences.
SDI Productions/E+ via Getty Images

Telling science's stories

Scientists, of course, can be science communicators, but everyone can tell science's stories. When we share information online about health, or talk to friends and about the weather, we contribute to information that circulates about science topics.

My son's pediatrician was a science communicator when she explained the vaccine schedule and ways to keep my son comfortable after receiving vaccines. I was a science communicator when I spoke to others about my decisions to fully vaccinate my son on the recommended schedule, and how he is now a healthy and happy 9-month-old.

When communicating about science topics, remember to borrow features from stories to strengthen your message. Think about all of a story's features – character, action, sequence, scope, storyteller and content – and how you might incorporate them into the topic. Everyone can find opportunities to strengthen their science communication, whether it's in their or in their everyday interactions with friends and family.The Conversation

Emma Frances Bloomfield, Associate Professor of Communication Studies, University of Nevada, Las Vegas

Advertisement

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Storytelling strategies make communication about science more compelling appeared first on theconversation.com

Advertisement
Continue Reading

News from the South

Trending