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, Congress can legislate accountability.

  • Instead of licensing companies to release advanced AI technologies, the could license auditors and push for companies to set up institutional review 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 comes next for Congress. The 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 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, 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 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 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 training. 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 .

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 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 benefits 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 week. 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

The perilous past and promising future of a toxic but nourishing crop

Published

on

theconversation.com – Stephen Wooding, Assistant Professor of Anthropology and Heritage Studies, of California, Merced – 2024-05-01 07:36:48

A grower shows off his lush cassava garden.

Stephen Wooding, CC BY-ND

Stephen Wooding, University of California, Merced

The three staple crops dominating modern diets – corn, rice and wheat – are familiar to Americans. However, fourth place is held by a dark horse: cassava.

Advertisement

While nearly unknown in temperate climates, cassava is a key source of nutrition throughout the tropics. It was domesticated 10,000 years ago, on the southern margin of the Amazon basin in Brazil, and spread from there throughout the region. With a scraggly stem a few meters tall, a handful of slim branches and modest, hand-shaped leaves, it doesn't look like anything special. Cassava's humble appearance, however, belies an impressive combination of productivity, toughness and diversity.

Five people sit in background with several piles of peeled and unpeeled cassava tubers

People preparing to cassava, with some peeled tubers in the foreground.

Philippe Giraud/Corbis Historical via Getty Images

Over the course of millennia, Indigenous peoples bred it from a weedy wild plant into a crop that stores prodigious quantities of starch in potatolike tubers, thrives in Amazonia's poor soils and is nearly invulnerable to pests.

Cassava's many assets would seem to make it the ideal crop. But there's a problem: Cassava is highly poisonous.

Advertisement

How can cassava be so toxic, yet still dominate diets in Amazonia? It's all down to Indigenous ingenuity. For the past 10 years, my collaborator, César Peña, and I have been studying cassava gardens on the Amazon River and its myriad tributaries in Peru. We have discovered scores of cassava varieties, growers using sophisticated breeding strategies to manage its toxicity, and elaborate methods for processing its dangerous yet nutritious products.

Long history of plant domestication

One of the most formidable challenges by early humans was getting enough to eat. Our ancient ancestors relied on hunting and gathering, catching prey on the run and collecting edible plants at every . They were astonishingly good at it. So good that their populations soared, surging out of humanity's birthplace in Africa 60,000 years ago.

Even so, there was room for improvement. Searching the landscape for food burns calories, the very resource being sought. This paradox forced a trade-off for the hunter-gatherers: burn calories searching for food or conserve calories by staying home. The trade-off was nearly insurmountable, but humans found a way.

A little more than 10,000 years ago, they cleared the hurdle with one of the most transformative innovations in history: plant and animal domestication. People discovered that when plants and animals were tamed, they no longer needed to be chased down. And they could be selectively bred, producing larger fruits and seeds and bulkier muscles to eat.

Advertisement

Cassava was the champion domesticated plant in the neotropics. After its initial domestication, it diffused through the region, reaching sites as far north as Panama within a few thousand years. Growing cassava didn't completely eliminate people's need to search the forest for food, but it lightened the load, providing a plentiful, reliable food supply close to home.

Today, almost every rural across the Amazon has a garden. Visit any household and you will find cassava roasting on the fire, being toasted into a chewy flatbread called casabe, fermenting into the beer called masato, and steaming in soups and stews. Before adopting cassava in these roles, though, people had to figure out how to deal with its toxicity.

Processing a poisonous plant

One of cassava's most important strengths, its pest resistance, is provided by a powerful defense system. The system relies on two chemicals produced by the plant, linamarin and linamarase.

These defensive chemicals are found inside cells throughout the cassava plant's leaves, stem and tubers, where they usually sit idle. However, when cassava's cells are damaged, by chewing or crushing, for instance, the linamarin and linamarase react, releasing a burst of noxious chemicals.

Advertisement

One of them is notorious: cyanide gas. The burst contains other nasty substances as well, compounds called nitriles and cyanohydrins. Large doses of them are lethal, and chronic exposures permanently damage the nervous system. Together, these poisons deter herbivores so well that cassava is nearly impervious to pests.

Nobody knows how people first cracked the problem, but ancient Amazonians devised a complex, multistep process of detoxification that transforms cassava from inedible to delicious.

two women in hats peeling and shredding tubers

Women grind the cassava's starchy tubers into shreds.

Stephanie Maze/Corbis Historical via Getty Images

It begins with grinding cassava's starchy roots on shredding boards studded with fish teeth, chips of rock or, most often today, a rough sheet of tin. Shredding mimics the chewing of pests, causing the release of the root's cyanide and cyanohydrins. But they drift away into the , not into the lungs and stomach like when they are eaten.

Advertisement

Next, the shredded cassava is placed in rinsing baskets where it is rinsed, squeezed by hand and drained repeatedly. The action of the releases more cyanide, nitriles and cyanohydrins, and squeezing rinses them away.

Finally, the resulting pulp can be dried, which detoxifies it even further, or cooked, which finishes the process using heat. These steps are so effective that they are still used throughout the Amazon today, thousands of years since they were first devised.

man standing next to large vat with fire beneath, under thatch roof

People cooking cassava in the traditional way in the 1970s.

Education Images/Universal Images Group via Getty Images

A powerhouse crop poised to spread

Amazonians' traditional methods of grinding, rinsing and cooking are a sophisticated and effective means of converting a poisonous plant into a meal. Yet, the Amazonians pushed their efforts even further, taming it into a true domesticated crop. In addition to inventing new methods for processing cassava, they began keeping track and selectively growing varieties with desirable characteristics, gradually producing a constellation of types used for different purposes.

Advertisement

In our travels, we have found more than 70 distinct cassava varieties that are highly diverse, physically and nutritionally. They include types ranging in toxicity, some of which need laborious shredding and rinsing and others that can be cooked as is, though none can be eaten raw. There are also types with different tuber sizes, growth rates, starch production and drought tolerance.

Their diversity is prized, and they are often given fanciful names. Just as American supermarkets stock apples called Fuji, Golden Delicious and Granny Smith, Amazonian gardens stock cassavas called bufeo (dolphin), arpón (harpoon), motelo (tortoise) and countless others. This creative breeding cemented cassava's place in Amazonian cultures and diets, ensuring its manageability and usefulness, just as the domestication of corn, rice and wheat cemented their places in cultures elsewhere.

While cassava has been ensconced in South and Central America for millennia, its story is far from over. In the age of climate change and mounting efforts toward sustainability, cassava is emerging as a possible world crop. Its durability and resilience make it easy to grow in variable environments, even when soils are poor, and its natural pest resistance reduces the need to protect it with industrial pesticides. In addition, while traditional Amazonian methods for detoxifying cassava can be slow, they are easy to replicate and speed up with modern machinery.

two workers in white coats, hair caps and gloves show off white clumps they are bagging

Workers package frozen cassava in bags at a Florida food processing plant.

Juan Silva/The Image Bank via Getty Images

Advertisement

Furthermore, the preference of Amazonian growers to maintain diverse types of cassava makes the Amazon a natural repository for genetic diversity. In modern hands, they can be bred to produce new types, fitting purposes beyond those in Amazonia itself. These advantages spurred the first export of cassava beyond South America in the 1500s, and its range quickly spanned tropical Africa and Asia. Today, production in nations such as Nigeria and Thailand far outpaces production in South America's biggest producer, Brazil. These successes are raising optimism that cassava can become an eco-friendly source of nutrition for populations globally.

While cassava isn't a familiar name in the U.S. just yet, it's well on its way. It has long flown under the radar in the form of tapioca, a cassava starch used in pudding and boba tea. It's also the shelves in the snack aisle in the form of cassava chips and the baking aisle in naturally gluten- flour. Raw cassava is an emerging presence, too, showing up under the names “yuca” and “manioc” in stores catering to Latin American, African and Asian populations.

Track some down and give it a try. Supermarket cassava is perfectly safe, and recipes abound. Cassava fritters, cassava fries, cassava cakes … cassava's possibilities are nearly endless.


This article was co-authored by César Rubén Peña.The Conversation

Stephen Wooding, Assistant Professor of Anthropology and Heritage Studies, University of California, Merced

Advertisement

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

Continue Reading

The Conversation

Sourdough under the microscope reveals microbes cultivated over generations

Published

on

theconversation.com – Daniel Veghte, Senior Research Associate Engineer, The Ohio – 2024-04-30 07:28:14

Microbes make a home among the starch grains of your sourdough starter.

Daniel Veghte, CC BY-SA

Daniel Veghte, The Ohio State University

Sourdough is the oldest kind of leavened bread in recorded history, and people have been eating it for thousands of years. The components of creating a sourdough starter are very simple – flour and . Mixing them produces a culture where yeast and bacteria ferment the sugars in flour, making byproducts that give sourdough its characteristic and smell. They are also what make it rise in the absence of other leavening agents.

Advertisement

My sourdough starter, affectionately deemed the “Fosters” starter, was passed down to me by my grandparents, who received it from my grandmother's college roommate. It has followed me throughout my academic career across the country, from undergrad in New Mexico to graduate school in Pennsylvania to postdoctoral work in Washington.

Currently, it resides in the Midwest, where I work at The Ohio State University as a senior research associate, collaborating with researchers to characterize samples in a wide variety of fields ranging from food science to material science.

As part of one of the microscopy courses I instruct at the university, I decided to take a closer look at the microbial community in my 's sourdough starter with the microscope I use in my day-to-day research.

Microscopy image of rod-shaped bacteria, elongated and spherical yeast, and globular starch grains

Each sourdough starter has a unique mix of microbes.

Daniel Veghte, CC BY-SA

Advertisement

Scanning electron microscopes

Scanning electron microscopy, or SEM, is a powerful tool that can image the surface of samples at the nanometer scale. For comparison, a human hair is between 10 to 150 micrometers, and SEM can observe features that are 10,000 times smaller.

Since SEM uses electrons instead of light for imaging, there are limitations to what can be imaged in the microscope. Samples must be electrically conductive and able to withstand the very low pressures in a vacuum. Low-pressure environments are generally unfavorable for microbes, since these conditions will cause the water in cells to evaporate, deforming their structure.

To prepare samples for SEM analysis, researchers use a method called critical point drying that carefully dries the sample to reduce unwanted artifacts and preserve fine details. The sample is then coated with a thin layer of iridium metal to make it conductive.

Round metal disk on a platform surrounded by a large cylindrical device

Scanning electron microscopes can image samples at the nanoscale level.

Daniel Veghte, CC BY-SA

Advertisement

Exploring a sourdough starter

Since sourdough starters are created from wild yeast and bacteria in the flour, it creates a favorable for many types of microbes to flourish. There can be more than 20 different species of yeast and 50 different species of bacteria in a sourdough starter. The most robust become the dominant species.

You can visually observe the microbial complexity of sourdough starter by imaging the different components that vary in size and morphology, yeast and bacteria. However, a full understanding of all the diversity present in the starter would require a complete gene sequencing.

The main component that gives the starter texture are starch grains from the flour. These grains, colored green in the image, are identifiable as relatively large globular structures approximately 8 micrometers in diameter.

Microscopy image of rod-shaped bacteria, elongated and spherical yeast, and globular starch grains

A false-colored scanning electron microscope image of a sourdough starter shows starch grains (green), yeast (red) and bacteria (blue).

Daniel Veghte, CC BY-SA

Advertisement

Giving rise to the starter is the yeast, colored red. As the yeast grows, it ferments sugars from the starch grains and releases carbon dioxide bubbles and alcohol as byproducts that make the dough rise. Yeast generally falls in the range of 2 to 10 micrometers in size and are round to elongated in shape. There are two distinct yeast types visible in this image, one that is nearly round, at the bottom left, and another that is elongated, at the top right.

Bacteria, colored blue, metabolize sugars and release byproducts such as lactic acid and acetic acid. These byproducts act as a preservative and are what give the starter its distinctive sour smell and taste. In this image, bacteria have pill-like shapes that are approximately 2 micrometers in size.

Now, the next time you eat sourdough bread or sourdough waffles – try them, they're delicious! – you can visualize the rich array of microorganisms that give each piece its distinctive flavor.The Conversation

Daniel Veghte, Senior Research Associate Engineer, The Ohio State University

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

Advertisement
Continue Reading

The Conversation

‘What is a fact?’ A humanities class prepares STEM students to be better scientists

Published

on

theconversation.com – Timothy Morton, Rita Shea Guffey Chair of English, Rice – 2024-04-30 07:29:12

A favorite class focuses on the tendency to see meaningful patterns where there aren't any, such as constellations of .

Yuga Kurita/Moment via Getty Images

Timothy Morton, Rice University

Text saying: Uncommon Courses, from The Conversation

Uncommon Courses is an occasional from U.S. highlighting unconventional approaches to teaching.

Advertisement

Title of course:

What Is a Fact?

What prompted the idea for the course?

With all the conspiracy theories floating around in 2020 when hit, I wanted to my learn to identify and deal with them. I was also concerned about political propaganda. And in my STEM-heavy school, I wanted to showcase what humanities scholars can do. So I created this class, which is distilled humanities for freshmen. Almost every student so far has been a science, technology, engineering and math major.

What does the course explore?

We start with a week called What Is Data? In Latin, “data” just means “things that are given.” Data can be in the form of measurements: “This bowlful of weighs x.” But data can also mean “it reminds me of my grandma.” How can you tell when something could be meaningful, or whether it's just nonsense?

A later class that students find especially interesting is on apophenia, the tendency to see patterns where there aren't any, like the man in the Moon, or constellations of stars.

Advertisement

chart illustrating dots of data, colored and connected in various ways as information, knowledge, insight, wisdom and conspiracy theory

Conspiracy theories connect a lot of dots, but that doesn't make them right.

Screenshot of a meme

Why is this course relevant now?

A fact is an interpretation of data. In physics class, you learn how to interpret physics data, find patterns, relate those patterns to other ones, and produce facts about them. If your argument hangs together logically, your interpretation can appear in the journal Nature Physics.

Humanities classes, however, prepare you to understand what facts are, period – whether they're based on biology or on the Bible, nutrition science or novels.

What's a critical lesson from the course?

One critical lesson is that many big conspiracy theories such as QAnon are about jumping to conclusions as quickly as possible. Being a good student and a good scholar means accepting that what you're examining might not be meaningful or might not indicate a pattern. What we're exploring here is how not to jump to conclusions. And this lesson applies as much to stuff in the real world as it does to lab work.

Advertisement

What materials does the course feature?

We watch YouTuber hbomberguy debunking global warming denialism. We read Kurt Gödel on how logical systems must always be flawed. We read poems and stories, introducing science majors to interpreting artistic data, a every bit as rigorous as interpreting scientific data.

What will the course prepare students to do?

Without the kinds of critical thinking this course teaches, scientists can be susceptible to propaganda and unable to share their ideas effectively, whether it's in the media or to their colleagues, friends and family.

Students learn to look at the world with fresh, skeptical eyes. They learn to identify illogical arguments and rhetorical strong-arm tactics. In the Middle Ages, humanities – grammar, logic, rhetoric – prepared you to do science. What Is a Fact? is like that, helping students see how collecting data and being skeptical don't stop once you've left the lab. A questioning, open-minded attitude is an essential skill.The Conversation

Timothy Morton, Rita Shea Guffey Chair of English, Rice University

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

Advertisement
Continue Reading

News from the South

Trending