 |
|
|
|||||
On December 11, 2015, OpenAI arrived on the scene with a bang. Announced on the penultimate day of the Conference on Neural Information Processing Systems, an academic confab held in Montreals Palais des Congrs by Elon Musk, Sam Altman, and others, the organization had been in the planning for months (an infamous July 2015 meeting at the Rosewood Sand Hill Hotel brought on board many of OpenAIs key early staffers). But when it went public with an announcement and blog post, the community reacted with surprise. This is just absolutely wonderful news, and I really feel like we are watching history in the making, wrote Sebastien Bubeck, then a researcher at Microsoft, and since October 2024, an OpenAI employee. The company was well-funded and professed to have clear goals: to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. Bubeck had little idea how prescient his words were. Even the wildest predictions of its founders on that day in 2015 likely couldnt have imagined how much ChatGPT would change the worldand OpenAIs fortunes. But now that its a decade old, the main question its investors, its employees, and all of us relying on its success to keep the stock market healthy are asking is: Where will it be in another 10-years time? Ten years ago, OpenAI started with a fairly legitimate scientific question and had a social conscious focus, says Catherine Flick, an AI ethicist at the University of Staffordshire. Flick points out that its founding form was a complicated nonprofit organization that was always going to be difficult to addressand caused plenty of consternation, including the2023 ousting of its CEO, Altman. But that founding ideal has changed significantly, she says. Now we have a for-profit company that has completely shared any responsibility for social benefit and has basically embraced that growth at all costs kind of mantra, says Flick. The reason? OpenAI is at the vanguard of the generative AI revolution, and theres money to be made. One key area that is likely to change OpenAI is the advent of superintelligence, a contested idea that the AI systems the company and its competitors are developing will at some point surpass human capabilities in every aspect. Those working closest to the AI models in frontier labs seem convinced of the idea that this will happenbut outsiders question whether thats simply a case of being too close to the Kool-Aid than superior knowledge as a result of seeing behind the scenes. Nevertheless, those at the top of the company are thinking about the future impact of AI a decade out. Earlier this year, OpenAI CEO Altman predicted that by 2035, college graduates if they still go to college at all, could very well be leaving on a mission to explore the solar system on a spaceship in some completely new, exciting, super well-paid, super interesting job. “OpenAI says that with today’s level of understanding, obviously nobody should deploy superintelligencebut also that their top priority is to build this, says Steven Adler, a fellow at the Roots of Progress Institute, who spent nearly four years between 2020 and 2024 at OpenAI, working in its safety research team. Its a concerning combination of beliefs, he adds. Adler hopes Open AIs plans for the future can remain independent and impartial from its for-profit interestssomething they are required to do Having overlapping membership of the for-profit and not-for-profit boards is a natural conflict, he believes. But more fundamentally, that will be challenging in part because of the competition that the company faces from other AI labs working at the frontier of the technology. “We all need to find ways to stop the AI industry’s race dynamicswhich OpenAI has long warned aboutfrom driving us off a cliff. Theres still some who think it could drive off that cliff. In 10-years, time, I fully expect OpenAI to have either completely imploded with all of its assets, sold off to some sort of private equity firm or similar, or have been snatched up by some other company and acquired in some way, Flick says. OpenAIs success will determine the AI industrys future, as well as that of the broader economy. Given its centrality to AI, OpenAIs success or failure and the rate of its process ultimately has major ramifications on the broader consumer internet and AI hyperscaler spaces, Ross Sandler, managing director and senior internet research analyst at Barclays, wrote in a recent research note. At present,OpenAI is sitting pretty, reckons Sandler, standing around six to 12 months ahead of its competitors in most areasthough the biggest firms like Google are starting to catch up with the release of its latest models like Gemini 3. Barclays estimates that by 2030, its revenue could be $200 billion, up from an estimated $13 billion this year, with around 44% of that coming from ChatGPT.Sandler also points out that OpenAI needs only convert a low single-digit proportion of its users in order to meet its revenue targets at present. That puts it well below other subscription apps, including Tinder, Spotify, and Duolingo. On one hand, Barclays research suggests that OpenAI is sitting pretty in its position at present. Weve found over the years that once habits are established, its hard for followers to dislodge the leader in a space, wrote Sandler. Yet Sandler also says that Google is the potentially huge spoiler in the midst for the next decade. For now, OpenAI is sitting pretty
Category:
E-Commerce
Large language models are quietly reshaping the way people write research papersand scientists are catching colleagues using AI to do their work.
Category:
E-Commerce
Generative artificial intelligence has become widely accepted as a tool that increases productivity. Yet the technology is far from mature. Large language models advance rapidly from one generation to the next, and experts can only speculate how AI will affect the workforce and peoples daily lives. As a materials scientist, I am interested in how materials and the technologies that derive from them affect society. AI is one example of a technology driving global changeparticularly through its demand for materials and rare minerals. But before AI evolved to its current level, two other technologies exemplified the process created by the demand for specialized materials: cars and smartphones. Often, the mass adoption of a new invention changes human behavior, which leads to new technologies and infrastructures reliant upon the invention. In turn, these new technologies and infrastructures require new or improved materialsand these often contain critical minerals: those minerals that are both essential to the technology and strain the supply chain. The unequal distribution of these minerals gives leverage to the nations that produce them. The resulting power shifts strain geopolitical relations and drive the search for new mineral sources. New technology nurtures the mining industry. The car and the development of suburbs At the beginning of the 20th century, only 5 out of 1,000 people owned a car, with annual production around a few thousand. Workers commuted on foot or by tram. Within a 2-mile radius, many people had all they needed: from groceries to hardware, from school to church, and from shoemakers to doctors. Then, in 1913, Henry Ford transformed the industry by inventing the assembly line. Now, a middle class family could afford a car: Mass production cut the price of the Model T from US$850 in 1908 to $360 in 1916. While the Great Depression dampened the broad adoption of the car, sales began to increase again after the end of World War II. With cars came more mobility, and many people moved farther away from work. In the 1940s and 1950s, a powerful highway lobby that included oil, automobile, and construction interests promoted federal highway and transportation policies, which increased automobile dependence. These policies helped change the landscape: Houses were spaced farther apart, and located farther away from the urban centers where many people worked. By the 1960s, two-thirds of American workers commuted by car, and the average commute had increased to 10 miles. Public policy and investment favored suburbs, which meant less investment in city centers. The resulting decay made living in downtown areas of many cities undesirable and triggered urban renewal projects. Long commutes added to pollution and expenses, which created a demand for lighter, more fuel-efficient cars. But building these required better materials. In 1970, the entire frame and body of a car was made from one steel type, but by 2017, 10 different, highly specialized steels constituted a vehicles lightweight form. Each steel contains different chemical elements, such as molybdenum and vanadium, which are mined only in a few countries. While the car supply chain was mostly domestic until the 1970s, the car industry today relies heavily on imports. This dependence has created tension with international trade partners, as reflected by higher tariffs on steel. The cellphone and American life The cellphone presents another example of a technology creating a demand for minerals and affecting foreign policy. In 1983, Motorola released the DynaTAC, the first commercial cellular phone. It was heavy, expensive, and its battery lasted for only half an hour, so few people had one. Then in 1996, Motorola introduced the flip phone, which was cheaper, lighter, and more convenient to use. The flip phone initiated the mass adoption of cellphones. However, it was still just a phone: Unlike todays smartphones, all it did was send and receive calls and texts. In 2007, Apple redefined communication with the iPhone, inventing the touchscreen and integrating an internet navigator. The phone became a digital hub for navigating, finding information, and building an online social identity. Before smartphones, mobile phones supplemented daily life. Now, they structure it. In 2000, fewer than half of American adults owned a cellphone, and nearly all who did used it only sporadically. In 2024, 98% of Americans over the age of 18 reported owning a cellphone, and over 90% owned a smartphone. Without the smartphone, most people cannot fulfill their daily tasks. Many individuals now experience nomophobia: They feel anxious without a cellphone. Around three-quarters of all stable elements are represented in the components of each smartphone. These elements are necessary for highly specialized materials that enable touchscreens, displays, batteries, speakers, microphones, and cameras. Many of these elements are essential for at least one function and have an unreliable spply chain, which makes them critical. Critical materials and AI Critical materials give leverage to countries that have a monopoly in mining and processing them. For example, China has gained increased power through its monopoly on rare earth elements. In April 2025, in response to U.S. tariffs, China stopped exporting rare earth magnets, which are used in cellphones. The geopolitical tensions that resulted demonstrate the power embodied in the control over critical minerals. The mass adoption of AI technology will likely change human behavior and bring forth new technologies, industries, and infrastructure on which the U.S. economy will depend. All of these technologies will require more optimized and specialized materials and create new material dependencies. By exacerbating material dependencies, AI could affect geopolitical relations and reorganize global power. America has rich deposits of many important minerals, but extraction of these minerals comes with challenges. Factors including slow and costly permitting, public opposition, environmental concerns, high investment costs, and an inadequate workforce all can prevent mining companies from accessing these resources. The mass adoption of AI is already adding pressure to overcome these factors and to increase responsible domestic mining. While the path from innovation to material dependence spanned a century for cars and a couple of decades for cellphones, the rapid advancement of large language models suggests that the scale will be measured in years for AI. The heat is already on. Peter Müllner is a distinguished professor in materials science and engineering at Boise State University. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||