Reporters are a skeptical bunch, so it was unusual to hear so many of them raving about their firsthand experience with Apple’s next Big Thing: the high-priced headset called Vision Pro, a device infused with totally virtual reality as well as augmented reality that projects digital images on top of real-world settings.

But after wearing the Vision Pro during a half-hour demonstration meticulously orchestrated by Apple, I joined the ranks of those blown away by all the impressive technology Apple has packed into the goggles-like headset. Still, that excitement was muted by a disquieting sense of having just passed through a gateway that eventually will lead society down another avenue of digital isolation.

THE POTENTIAL UPSIDES

But first the good stuff: Vision Pro is a highly sophisticated device that is fairly easy to set up and incredibly intuitive to use. The setup requires using an iPhone to automatically take some assessments of your eyes and ears. If you wear prescription glasses (I wear contacts) some additional calibration will be needed, but Apple promises that won’t be complicated.

Once that’s all done, you will quickly find that putting on the Vision Pro is also simple, thanks to a knob on the side that makes it easy to ensure a the headset fits comfortably. And unlike other headsets, the Vision Pro isn’t an awkward-looking piece of nerdware, although the goggles aren’t exactly chic, despite looking a bit like something you might see people wearing on a ski slope, jet fighter or race car.

Controlling the Vision Pro is astoundingly easy. Users just press a button above the right goggle to pull up a virtual screen of apps, including familiar standbys for photos, messaging, phone calls, video streaming and web browsing. Opening an app just requires looking straight at it, then pinching a thumb and finger together. The same app can be closed with a finger pinch or can be moved to the side by holding two fingers together and moving them in the direction where you want to place it.

Not surprisingly, Apple’s well-curated demonstration cast the Vision Pro in the best-possible light. The headset clearly seems like it could be quite popular for business purposes, improving productivity, collaboration and video conferencing, especially in an era when more work is being done remotely.

Without causing the disorienting effects common in other virtual-reality headsets, the Vision Pro can immerse you in stunning visuals, 3-D displays of faraway places. It can insert you into videos of past memories recorded with one of the device’s 12 cameras (the demo included heartwarming scenes of a child’s birthday party and a campfire scene). It can make watching a 3-D movie, such as the latest Avatar film, feel like you are sitting in an IMAX theater while relaxing on your own couch. It can thrust you into surreal moments (at one point, I watched in wonder as a butterfly first shown in a virtual screen depicting a prehistoric era seemingly fluttered across the room and landed in my outstretched hand as I sat on a couch).

And the demo featured just enough glimpses of the way sporting events appear through the goggles to realize that the powers that be in professional and collegiate football, basketball, baseball and hockey are bound to find ways to incorporate the technology into subscription services that make viewers feel like they are sitting in the front row.

To Apple’s credit, the Vision Pro is also designed in a way that allows users to still see those around them, if they so choose.

THE POTENTIAL DOWNSIDES

My mixed feelings about Apple’s first foray into mixed reality ironically stems from just how well-designed the Vision Pro is by a company that has been behind this sort of game-changing technology on numerous occasions during the past 40 years, ranging from the Macintosh computer to the iPhone.

It feels like this may be another instance in which Apple has accomplished something that has eluded other tech companies by cracking the code to make both virtual- and augmented-reality more compelling and less disorienting than a variety of other ho-hum headsets have done over the past decade or so.

The only reason the Vision Pro isn’t going to be an immediate sensation is its cost. When it hits the U.S. market early next year, it will sell for $3,500, which makes it probable it will start out as a luxury item unaffordable to most households—especially because the headset isn’t going to supplant the need to buy a new iPhone or smartphone running on Android every few years.

The most likely scenario is that Vision Pro in some ways is Apple’s testbed for mixed reality that will encourage the development of more apps especially designed to take advantage of the technology. The next ripple effect will be an array of other products equipped with similarly compelling technology at lower price points that stand a better chance sucking more people—including children—into a realm that threatens to deepen screen addictions to the detriment of real-world interactions among humans.

© 2023 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed without permission.

A team including researchers from the National Institute of Information and Communications Technology; Sumitomo Osaka Cement Co., Ltd.; Nagoya Institute of Technology; and Waseda University has jointly developed the world’s first system for the transparent relay, routing, and switching of high-speed terahertz-wave signals to different locations.

Direct conversion of a 32-Gb/s terahertz-wave signal in the 285-GHz band to an optical fiber and its transparent relay and switching to different access points in ultrashort time periods were successfully demonstrated.

The key technologies include a newly developed low-loss optical modulator for the direct conversion of terahertz-wave signals to optical signals and an adaptive fiber-wireless technology for the ultrafast switching of terahertz signals. The developed system overcomes the disadvantages of radio communications in the terahertz band, such as high free-space loss, weak penetration, and limited communication coverage, paving the way for the deployment of terahertz communications in beyond 5G and 6G networks.

The results of this demonstration were published as a post-deadline paper at the 2023 International Conference on Optical Fiber Communications (OFC 2023).

Terahertz background

Radio frequencies in the terahertz band are promising candidates for ultrahigh-data-rate communications in beyond 5G and 6G networks. A 160-GHz slot in the 275–450 GHz band was recently opened for mobile and fixed services. However, high free-space loss and weak penetration remain as bottlenecks, making it difficult to transmit signals over long distances, such as from outdoors to indoors or in environments with obstacles.

Transparent relay and routing of terahertz signals between different locations are crucial to overcoming these disadvantages and expanding communication coverage. However, these functions cannot be realized using current electronic technologies. In addition, the narrow beamwidth of terahertz signals makes it difficult to achieve uninterrupted communication when users are moving. Terahertz-signal switching between different directions and locations is crucial for maintaining communication with end users.

However, this critical issue has not yet been addressed using electronic or photonic technologies. It is also important to turn on and off the emission of terahertz signals at appropriate intervals to save energy and reduce interference.

Study achievements

In this study, the team demonstrated the first system for the transparent relay, routing, and switching of terahertz signals in the 285-GHz band utilizing two key technologies: (i) a newly developed low-loss optical modulator and (ii) an adaptive fiber–wireless technology using an ultrafast wavelength-tunable laser. In the system, terahertz signals are received and directly converted into optical signals using terahertz–optical conversion devices with low-loss optical modulators.

Lightwave signals from tunable lasers are input to the modulators, and wavelength routers are used to route the signals to different access points where specific wavelengths are assigned. At the access points, the modulated optical signals are converted back into terahertz signals using optical-terahertz converters. Terahertz signals can be switched to different access points by switching the wavelengths of the tunable lasers.

The tunable lasers can be independently controlled, and the number of access points that can simultaneously generate terahertz signals equals the number of active tunable lasers. Using the technologies developed in this study, the team successfully demonstrated the transparent relay and switching of terahertz signals in the 285-GHz band for the first time and achieved a transmission capacity of 32 Gb/s using a 4-quadrature amplitude modulation (QAM) orthogonal frequency division multiplexing (OFDM) signal.

The possibility of switching the terahertz-wave signals in less than 10 μs was evaluated, confirming that uninterrupted communication can be attained in the terahertz band.

The system consists of the following key element technologies:

• Direct conversion of terahertz signals to optical signals using a newly developed low-loss optical modulator. The team achieved this by performing Ti diffusion on the x-cut lithium niobate (LiNbO₃ thickness: ≤ 100 µm) in the low dielectric constant layer for operation up to 330 GHz.
• Ultrafast terahertz-signal switching by controlling the wavelengths of tunable lasers to route and distribute terahertz signals to different locations
• 4-QAM OFDM signal transmission

By transparently relaying and distributing terahertz signals to different locations, high free-space and penetration losses of radio signals in the terahertz band can be overcome, and communication coverage can be significantly extended. In addition, by promptly routing and switching the signals to different directions/locations, communication can be maintained even when obstacles occur and/or users are moving.

Furthermore, by turning on and off the emission of terahertz-wave signals from access points at appropriate intervals, energy savings and interference reduction can be achieved. These features render the proposed system a promising solution for overcoming the bottlenecks of terahertz-wave communications and paving the way for its deployment in beyond 5G and 6G networks.

In the future, the researchers will study the terahertz-optical conversion device and fiber-wireless technology developed in this study to further increase the radio frequency and transmission capacity. In addition, they will promote international standardization and social implementation activities related to fiber-wireless and terahertz-wave communication systems.

Provided by
National Institute of Information and Communications Technology (NICT)

A new report by the International Energy Agency (IEA), the International Renewable Energy Agency (IRENA), the United Nations Statistics Division (UNSD), the World Bank, and the World Health Organization (WHO), released today, finds that the world is not on track to achieve the Sustainable Development Goal (SDG) 7 for energy by 2030.

This year marks the halfway point for achieving SDGs by 2030. SDG 7 is to ensure access to affordable, reliable, sustainable and modern energy. The goal includes reaching universal access to electricity and clean cooking, doubling historic levels of efficiency improvements, and substantially increasing the share of renewables in the global energy mix. Attaining this goal will have a deep impact on people’s health and well-being, helping to protect them from environmental and social risks such as air pollution, and expanding access to primary health care and services.

The 2023 edition of Tracking SDG 7: The Energy Progress Report warns that current efforts are not enough to achieve the SDG 7 on time. There has been some progress on specific elements of the SDG 7 agenda—for example, the increased rate of using renewables in the power sector—but progress is insufficient to reach the targets set forth in the SDGs.

The global energy crisis is expected to stimulate the deployment of renewables and improve energy efficiency with several government policies pointing to increasing investment. However, IRENA estimates show that international public financial flows in support of clean energy in low- and middle-income countries have been decreasing since before the COVID-19 pandemic and funding is limited to a small number of countries. To meet SDG 7 targets and to ensure that people fully benefit from the socioeconomic gains of the shift to sustainable energy, it is necessary to structurally reform international public finance and define new opportunities to unlock investments.

The report also finds that mounting debt and rising energy prices are worsening the outlook for reaching universal access to clean cooking and electricity. Current projections estimate that 1.9 billion people will be without clean cooking and 660 million without electricity access in 2030 if we do not take further action and continue with current efforts.

These gaps will negatively impact the health of our most vulnerable populations and accelerate climate change. According to WHO, 3.2 million people die each year from illness caused by the use of polluting fuels and technologies, which increase exposure to toxic levels of household air pollution.

Key findings of the report

• In 2010, 84% of the world’s population had access to electricity. This increased to 91% in 2021, meaning more than a billion people gained access over that period. However, the growth pace of access slowed in 2019–2021 compared to previous years. Rural electrification efforts contributed to this progress, but a large gap within urban areas remains.
• In 2021, 567 million people in sub-Saharan Africa did not have access to electricity, accounting for more than 80% of the global population without access. The access deficit in the region stayed almost the same as in 2010.
• The world remains off track to achieve universal access to clean cooking by 2030. Up to 2.3 billion people still use polluting fuels and technologies for cooking, largely in sub-Saharan Africa and Asia. The use of traditional biomass also means households spend up to 40 hours a week gathering firewood and cooking, which prohibits women from pursuing employment or participating in local decision-making bodies and children from going to school.
• According to the 2019 WHO estimates, 3.2 million premature deaths each year were attributable to household air pollution created by using polluting fuels and technologies for cooking.
• Renewable electricity use in global consumption has grown from 26.3% in 2019 to 28.2% in 2020, the largest single-year increase since the start of tracking progress for the SDGs.
• Efforts to increase renewables’ share in heating and transport, which represent more than three quarters of global energy consumption, remain off target to achieve 1.5oC climate objectives.
• Energy intensity—the measure of how much energy the global economy uses per dollar of GDP—improved from 2010 to 2020 by 1.8% annually. This is higher than the 1.2% improvement from the previous decades.
• However, the rate of energy intensity improvement has slowed in recent years and dropped to 0.6% in 2020. This makes it the worst year for energy intensity improvement since the global financial crisis, albeit largely due to pandemic-related restrictions, which may indicate only a temporary setback. Annual improvements through 2030 must now average 3.4% to meet the SDG target 7.3.
• International public financial flows in support of clean energy in developing countries stand at US $10.8 billion in 2021, 35% less than the 2010–2019 average and only about 40% of the 2017 peak of US $26.4 billion. In 2021, 19 countries received 80% of the commitments.

The report will be presented to top decision-makers at a special launch event on 11 July 2023 at the High-Level Political Forum (HLPF) on Sustainable Development, ahead of the second SDG Summit in September 2023 in New York. The authors urge the international community and policymakers to safeguard the gains made toward achieving SDG 7, to advance structural reforms, and to maintain a strategic focus on the vulnerable countries needing the most support.

“The energy crisis sparked by Russia’s invasion of Ukraine continues to have a profound impact on people all around the world. High energy prices have hit the most vulnerable hard, particularly those in developing economies. While the clean energy transition is moving faster than many think, there is still a great deal of work needed to deliver sustainable, secure and affordable access to modern energy services for the billions of people who live without it. Successful energy transitions rely on effective policies and technological innovation combined with large-scale mobilization of investment capital. The international community must leverage all these tools to meet the Sustainable Development Goals by the end of this decade,” said Fatih Birol, Executive Director, International Energy Agency.

“Cost-competitive renewable energy has yet again demonstrated remarkable resilience, but the poorest in the world are still largely unable to fully benefit from it. To realize SDG7 without compromising climate goals, we must bring about systemic change in the way international cooperation works. It is crucial that multilateral financial institutions direct financial flows more equitably around the world to support renewables deployment and related physical infrastructure development,” noted Francesco La Camera, Director-General, International Renewable Energy Agency.

“Despite advances towards sustainable energy targets at the mid-point of Agenda 2030, Goal 7 seems harder to reach than it was in 2015 and scaled-up action is necessary if we are to ensure access to affordable, reliable, sustainable and modern energy for all by 2030. Access to electricity and clean cooking still display great regional disparities and should be the focus of action to ensure that no one is left behind. Investment needs to reach the least-developed countries and sub-Saharan Africa to ensure more equitable progress toward Goal 7,” remarked Stefan Schweinfest, United Nations Statistics Division.

“Despite a recent slowdown in the global pace of electrification, the number of people without electricity almost halved over the past decade, from 1.1 billion in 2010 down to 675 million in 2021. Nonetheless, additional efforts and measures must urgently be put in place to ensure that the poorest and hardest-to-reach people are not left behind. To reach universal access by 2030, the development community must scale up clean energy investments and policy support,” added Guangzhe Chen, Vice President for Infrastructure, World Bank.

“We must protect the next generation by acting now. Investing in clean and renewable solutions to support universal energy access is how we can make real change. Clean cooking technologies in homes and reliable electricity in health-care facilities can play a crucial role in protecting the health of our most vulnerable populations,” concluded Dr. Tedros Adhanom Ghebreyesus, WHO Director-General, World Health Organization.

Provided by
IEA

A simple sponge has improved how robots grasp, scientists from the University of Bristol have found.   

This easy-to-make sponge-jamming device can help stiff robots handle delicate items carefully by mimicking the nuanced touch, or variable stiffness, of a human.

Robots can skip, jump and do somersaults, but they’re too rigid to hold an egg easily. Variable-stiffness devices are potential solutions for contact compliance on hard robots to reduce damage, or for improving the load capacity of soft robots.

This study, published at the IEEE International Conference on Robotics and Automation (ICRA) 2023, shows that variable stiffness can be achieved by a silicone sponge.

Lead author Tianqi Yue from Bristol’s Department of Engineering Mathematics explained, “Stiffness, also known as softness, is important in contact scenarios.”

“Robotic arms are too rigid so they cannot make such a soft human-like grasp on delicate objects, for example, an egg.”

“What makes humans different from robotic arms is that we have soft tissues enclosing rigid bones, which act as a natural mitigating mechanism.”

“In this paper, we managed to develop a soft device with variable stiffness, to be mounted on the end robotic arm for making the robot-object contact safe.”

Silicone sponge is a cheap and easy-to-fabricate material. It is a porous elastomer just like the cleaning sponge used in everyday tasks.

By squeezing the sponge, the sponge stiffens which is why it can be transformed into a variable-stiffness device.

This device could be used in industrial robots in scenarios including gripping jellies, eggs and other fragile substances. It can also be used in service robots to make human-robot interaction safer.

Mr. Yue added, “We managed to use a sponge to make a cheap and nimble but effective device that can help robots achieve soft contact with objects. The great potential comes from its low cost and light weight.

“We believe this silicone-sponge based variable-stiffness device will provide a novel solution in industry and healthcare, for example, tunable-stiffness requirement on robotic polishing and ultrasound imaging.”

The team will now look at making the device achieve variable stiffness in multiple directions, including rotation.

Artificial intelligence may seem similar to a calculator but the relationship of humans with the former is not as serene as with the latter. We trust the results of computers, even if we don’t know exactly how it can arrive at the result of a complex operation in a short time, while the relationship with artificial intelligence (AI) generates discomfort in people. Why? The reason is that machines never stop learning and the more they perform new and unexpected tasks, hitherto entrusted to human intelligence, the more users distrust them, because they do not like to find their own prerogatives embodied in machines. This is what creates discomfort.

There is a fear of being overtaken by some super-intelligence and, staying in the economic sphere, of being replaced in the workplace by machines. And yet, we are immersed every day in algorithms and we deal with programs such as bots in our roles of telephone customers or bank savers. It is believed that bots are behind about 50% of internet traffic, that 40% of Wikipedia edits are the work of digital agents, without forgetting the presence on social networks of many accounts created by automated platforms and, last but not least, the boom in popular curiosity for ChatGPT, capable of producing written texts based on conversations with users.

In reality, “Algorithms work and evolve precisely because they no longer try to be intelligent. If anything, they can be seen as capable of communicating creatively and for informational purposes, but they are cannot be considered intelligent,” explains Elena Esposito, Full Professor of Sociology of Cultural and Communicative Processes at Bielefeld and Bologna. She recently put out the book “Comunicazione artificiale. Come gli algoritmi producono intelligenza sociale” (“Artificial communication. How algorithms produce social intelligence,” Bocconi University Press, 2022).

“In fact, the purpose with which algorithms are programmed is not to understand the data provided by our online behavior. The intention is to identify correlations between data and process them so that they are informative for users,” says Esposito, a student of Niklas Luhmann, who not by chance proposes to move from the definition of artificial intelligence to that of artificial communication, positing a new theoretical model to reiterate that the interlocutor with whom we interact is not a human being, but an algorithm. We need new rules and habits of behavior which, given the AI’s multiple areas of application, must be promoted by national and supranational institutions, as well as families and individuals alike, according to their respective realms of pertinence.

How can talking about artificial communication reduce the discomfort felt towards AI?

First, because I hypothesize that the analogy between the performance of algorithms and human intelligence, which generates this discomfort, is misleading. Furthermore, because it allows for the emergence of new insights on the challenges and paradoxes that recent technologies pose. The numerous positive aspects of algorithmic intelligence remain, from the availability of more information to the higher speed with which to find it, passing through the cost-effectiveness of the process. But we can also ask ourselves how the growing intervention of AI affects, for example, our conception of the public sphere and the maintenance of social cohesion, considering the progressive customization of the information and services offered to each individual person, without he or she having even asked for them.

This creates a bubble that is difficult to get out of. It becomes more difficult to realize that there may be something different from what you already know and decide whether you want to find it out or not. In other words, the individual no longer knows what others know and that common ground of shared information that makes everyone feel part of the societal whole decays. But social cohesion and markets themselves vitally depend on that common ground of shared. In addition, one may wonder what are the effects of the different versions of AI in specific fields such as education or, finally, how our perception of the relationship between reality and fiction changes. More and more often, in fact, we can intervene not only on reality but also on fiction, which is no longer the unalterable fiction of commercial movies or novel, by but an area with which we can interact and during the course of the story, such as for example it happens in video games.

The growing presence of AI in our lives changes our faculties? For example, in knowing how to remember and what to forget?

Let’s say that there is a new relationship between people and oblivion. In the past we mostly committed ourselves to remembering things and forgetting proceeded by itself, it intervened spontaneously to select the information that shouldn’t last over time. Now the difficulty is reversed and lies in remembering not to remember; you need to try harder to forget as all memories and information are preserved online. We can therefore reason on the final paradox we arrive at: to forget memories it is better to multiply them, to make one climb from the first place to the eleventh in the results of a search engine, since we know that people tend to read only the first few results.

Is it believable or illusory that machines can predict our future?

The future will forever be unwritten because it depends on human behavior, which is constantly changing. The future remains open, even if we can remember that AI offers a series of new tools to deal with the uncertainty of the future. However, if so far we have tried to anticipate it by relying on the calculation of probabilities, now the algorithms try to identify correlations between various possible configurations in the big data sets. And the correlations highlighted are not necessarily the most likely, even if the algorithms are partly based on probabilistic structures.

The result of trying to predict the future is that algorithms produce indications about the future that are obscure to humans, since they are unable to understand how they were generated. These are predictions that end up recalling the divinatory practices of the ancient world, with their sibylline and cryptic responses. Precisely the practices from which science had to move away from.

Big data, machine learning and bots, there are some of the terms that describe the hi-tech scenario in which we are immersed. The implications of these technologies currently unfolding in our lives are vast. Thus questions arise such as “Will we able to control something that we do not fully understand?” or “Aren’t the machines just getting too smart?”

Provided by
Bocconi University

Smart digital image sensors that can perform visual perception capabilities, such as scene recognition, are the result of recent research at KAUST.

Exploiting the Nobel-Prize-winning technology of the charge-coupled device (CCD) image sensors found in early digital cameras, Dayanand Kumar, Nazek El-Atab and co-workers have adapted and enhanced the CCD’s core structure to create light-sensitive memory devices that can be programmed by light. In particular, the research team embedded the two-dimensional material MoS₂ into a semiconductor capacitor (MOSCAP) structure that underpins the charge-storing pixels of a CCD sensor.

The resulting Al/Al₂O₃/MoS₂/Al₂O₃/Si MOSCAP structures function as a charge-trapping “in-memory” sensor that is sensitive to visible light and can be programmed optically and erased electrically. The work was published in the journal Light: Science & Applications.

“The in-memory light sensors are smart multifunctional memory devices that can perform the roles of multiple—traditionally discrete—devices at once, including optical sensing, storage and computation,” explained El-Atab.

“Our long-term goal is to be able to demonstrate in-memory sensors that can detect different stimuli and compute,” El-Atab explains. “This overcomes the memory wall and allows for faster and more real-time data analysis using reduced power consumption, which is a requirement in many futuristic and state-of-the-art applications such as Internet of Things, autonomous cars and artificial intelligence, among others.”

Experiments with light with a wavelength anywhere in the blue to the red spectral region indicate that a photo-generated charge can be trapped or stored with an extremely long-lived retention time. The resulting “memory window” voltage of >2V can be stored for up to 10 years prior to being electrically erased by applying a +/-6V signal. Moreover, it can be operated for many millions of cycles.

The ultimate aim of the research is to create a single optoelectronic device that can perform optical sensing and storage with computing capabilities.

By combining their MoS₂ MOSCAP structure with a neural network, the team showed that it was possible to perform simple binary image recognition, successfully distinguishing between images of either a dog or an automobile, with an accuracy of 91%. Each image was 32×32 pixels in size, and only the blue information from the images was extracted since that corresponds to the device’s peak sensitivity.

The team would now like to further develop the level of optical control. “Current memory devices can be programmed optically but require erasing electrically,” commented Kumar. “In the future, we would like to explore in-memory optical sensors that can be fully optically operated.”

In related work published in Advanced Materials, also on the topic of using optoelectronics to perform tasks for artificial visual perception, the team has explored the use of black phosphorus to create an optoelectronic memristive synapse that mimics the brain’s neurons for neuromorphic computing applications.

Their multilayer device consists of a thin layer of black phosphorus and hafnium oxide that is sandwiched between a lower layer of platinum and an upper layer of copper. It operates as an optoelectronic memristor—a resistor that can have its electrical resistance programmed by visible light.

Experiments indicate that it offers highly stable synaptic features, such as long-term potentiation (a long-lasting increase in signal output), long-term depression (a long-lasting decrease in signal output) and short-term plasticity (change in response over time), which are all important neuronal behaviors.

The team constructed a 6×6 synaptic array from the devices, and in the future they hope that larger arrays could help realize a biomimetic retina. Importantly, the devices can be fabricated cost effectively by solution processing and are flexible with stable operation with a bend radius of 1 centimeter, offering possibilities for wearable applications.

Artificial intelligence poses an “existential risk” to humanity, a key innovator warned during a visit to the United Arab Emirates on Tuesday, suggesting an international agency like the International Atomic Energy Agency oversee the ground-breaking technology.

OpenAI CEO Sam Altman is on a global tour to discuss artificial intelligence.

“The challenge that the world has is how we’re going to manage those risks and make sure we still get to enjoy those tremendous benefits,” said Altman, 38. “No one wants to destroy the world.”

OpenAI’s ChatGPT, a popular chatbot, has grabbed the world’s attention as it offers essay-like answers to prompts from users. Microsoft has invested some $1 billion in OpenAI.

ChatGPT’s success, offering a glimpse into the way that artificial intelligence could change the way that humans work and learn, has sparked concerns as well. Hundreds of industry leaders, including Altman, have signed a letter in May that warns “mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”

Altman made a point to reference the IAEA, the United Nations nuclear watchdog, as an example of how the world came together to oversee nuclear power. That agency was created in the years after the U.S. dropping atom bombs on Japan at the end of World War II.

“Let’s make sure we come together as a globe—and I hope this place can play a real role in this,” Altman said. “We talk about the IAEA as a model where the world has said ‘OK, very dangerous technology, let’s all put some guard rails.’ And I think we can do both.

“I think in this case, it’s a nuanced message ’cause it’s saying it’s not that dangerous today but it can get dangerous fast. But we can thread that needle.”

Lawmakers around the world also are examining artificial intelligence. The 27-nation European Union is pursuing an AI Law that could become the de facto global standard for artificial intelligence. Altman told the U.S. Congress in May that government intervention will be critical to governing the risks that come with AI.

But the UAE, an autocratic federation of seven hereditarily ruled sheikhdoms, offers the flip side of the risks of AI. Speech remains tightly controlled. Rights groups warn the UAE and other states across the Persian Gulf regularly use spying software to monitor activists, journalists and others. Those restrictions affect the flow of accurate information—the same details AI programs like ChatGPT rely on as machine-learning systems to provide their answers for users.

Among speakers opening for Altman at the event at the Abu Dhabi Global Market was Andrew Jackson, the CEO of the Inception Institute of AI, which is described as a company of G42.

G42 is tied to Abu Dhabi’s powerful national security adviser and deputy ruler Sheikh Tahnoun bin Zayed Al Nahyan. G42’s CEO is Peng Xiao, who for years ran Pegasus, a subsidiary of DarkMatter, an Emirati security firm under scrutiny for hiring former CIA and NSA staffers, as well as others from Israel. G42 also owns a video and voice calling app that reportedly was a spying tool for the Emirati government.

In his remarks, Jackson described himself as representing “the Abu Dhabi and UAE AI ecosystem.”

“We are a political powerhouse and we will be central to AI regulation globally,” he said.

© 2023 The Associated Press. All rights reserved. This material may not be published, broadcast, rewritten or redistributed without permission.

Music streaming app Deezer said Tuesday it was launching a tool to detect and tag songs with AI-generated vocal clones in a bid to protect the revenues of the real artists.

Artificial Intelligence tools have recently allowed people to recreate the sound of famous artist vocals, from The Beatles to Oasis.

“Heart on a Sleeve”, a track featuring AI-generated copycats of Drake and The Weeknd, racked up millions of hits on TikTok and other platforms.

Music companies fear this could eat into future earnings since there is currently no way to copyright the sound of someone’s voice.

But Deezer said it plans to identify AI-generated music, focusing initially on those which recreate the voices of existing artists.

The French company said it aimed to create a system to tag these tunes in the app, which can alert artists, labels and users to this “fraudulent activity”.

“Our goal is to weed out illegal and fraudulent content, increase transparency, and develop a new remuneration system where professional artists are rewarded for creating valuable content,” said Deezer CEO Jeronimo Folgueira in a statement.

He said more than 100,000 new tracks are uploaded to Deezer’s site daily, making it increasingly important to distinguish between human- and machine-generated music.

“AI can be used to create new incredible content and I believe there are massive benefits of using generative AI,” he added. “But we need to ensure it’s done in a responsible way.”

© 2023 AFP

Taiwanese semiconductor giant TSMC said Tuesday its production of ever-smaller microchips will remain on the island, hopeful that the critical industry will have a “stabilizing effect on global geopolitical conflicts”.

Taiwan Semiconductor Manufacturing Company—whose clients include Apple and Intel—controls more than half the world’s output of silicon wafers, used in everything from smartphones to cars and missiles.

Recent years have seen it navigating geopolitical tussles between the United States and China, with the countries facing off over a range of issues including technology, trade, and Taiwan—the primary manufacturing base of the world’s semiconductors.

Alarm has spiked in recent months over the future of the semiconductor industry—the lifeblood of the global economy—if Taiwan were to be invaded by China, which considers the self-ruled island its own territory.

“I can assure everyone that for the three-nanometre, two-nanometre (chips), we will manufacture here in Taiwan,” said TSMC CEO CC Wei on Tuesday during a shareholder meeting in the northern city of Hsinchu.

TSMC’s three-nanometre and two-nanometre chips are expected to be among the world’s most advanced technologies—to be used in high-performing devices as the global market pivots towards products deploying artificial intelligence, like ChatGPT.

Wei also said that an even smaller product—the 1.4-nanometre chip—will be manufactured “almost entirely in Taiwan”.

China has ramped up military and political pressures on Taiwan, strongly protesting against any diplomatic engagement with the island.

In April, Beijing conducted three days of military exercises simulating a blockade of the island in response to US House Speaker Kevin McCarthy meeting Taiwanese President Tsai Ing-wen in California.

On Tuesday, TSMC sought to project an image of calm.

“I think Taiwan’s semiconductor industry plays a stabilizing role amid global geopolitical tensions. Whether it’s China or the United States… they all hope that TSMC is around,” said executive chairman Mark Liu.

“I hope Taiwan’s semiconductor industry can do well to have a stabilizing effect on global geopolitical conflict.”

The US-China trade conflict combined with the impacts of the Russia-Ukraine war and climate change have led to lowered expectations for the semiconductor market, said Liu, though TSMC’s performance in the second half of the year “will be better than the first half”.

He added TSMC is still negotiating with German authorities over a factory in Dresden, though they are concerned about the talent pool there and the country’s supply chain systems.

“There are some gaps indeed, but… the local government promises to build them up in the short-term,” said Liu. “We feel rather good about it.”

© 2023 AFP

The use of mobile technologies to collect and analyze individuals’ location information has produced massive amounts of consumer location data, giving rise to an elaborate multi-billion-dollar system in which consumers can share personal data in exchange for economic benefits. But privacy risks prevail.

In a new study, researchers have used machine learning to create and test a framework that quantifies personalized privacy risks; performs personalized data obfuscation; and accommodates a variety of risks, utilities, and acceptable levels of risk-utility tradeoff. The framework outperformed prior models, significantly reducing consumers’ privacy risk while preserving advertisers’ utility.

The study was conducted by researchers at Carnegie Mellon University (CMU), the University of Virginia, and New York University. It is published in Information Systems Research.

“The global market for location analytics alone is projected to reach $25.5 billion by 2027,” notes Beibei Li, associate professor of IT and management at CMU’s Heinz College, who co-authored the study. “As industries increasingly unleash the power of location big data, our study offers a much-needed framework to balance privacy risks and data utilities, and to sustain a secure and self-governing multi-billion-dollar location ecosystem.”

Massive volumes of mobile location data are being generated daily through smartphone location-based services (e.g., navigation, ride share, food delivery services). Such data track consumers’ behavior—where they eat and shop, what products they buy—to enable applications of commercial value (e.g., restaurant recommendations, location-based advertising, market research). Advertisers who gain access to location data through data aggregators can predict consumers’ next location with 25% success and next activity and timing with 26% success.

But there are considerable risks to consumers of sharing location data, which includes personally identifiable information like names and home addresses. Some advertisers may carry out malicious acts using the data, usually for short-term revenue gains. Therefore, data aggregators need a personalized and flexible framework to balance diverse types of risks and utilities for different kinds of consumers and advertisers.

In this study, researchers developed a machine learning-based framework that quantifies individual consumers’ privacy risk, quantifies advertisers’ utility, and features a personalized and flexible obfuscation scheme. The scheme suppresses a subset of locations visited by a consumer based on his or her personalized suppression parameter proportional to the individual’s risk level; it also accommodates different types and different acceptable levels of risks and utilities.

To test their framework, researchers partnered with a leading data aggregator that integrates location data across more than 400 commonly used mobile apps (e.g., news, weather, maps, fitness) from a quarter of the U.S. population who are in compliance with privacy regulations. The data, collected in five weeks from September to October 2018, are representative of the U.S. population and the sample analyzed covers a major U.S. metropolitan area. Researchers validated the framework on a million trajectories (where and when consumers move) generated by 40,000 consumers in a major U.S. metropolitan area.

The study’s framework accounts for distinct characteristics of individual-level location data, and outperforms multiple benchmark methods from recent studies, according to the authors.

Using the proposed framework, the authors say, a data aggregator can effectively curtail a potential invasion of consumer privacy by performing personalized data obfuscation without sacrificing the utility of the obfuscated data to an advertiser. The aggregator may also fulfill personalized and diverse demands from both consumers and advertisers by flexibly accommodating multiple types of risks and utilities, as well as a wide array of acceptable levels of a specific risk, utility, and risk-utility tradeoff.

“Location-based marketing is rapidly becoming a primary venue for planning marketing campaigns and targeting consumers, enriching both traditional and digital marketing strategies,” explains Meghanath Macha, a graduate of CMU’s Heinz College, who led the study. “Our framework fills a critical void and offers an important tool for the privacy-aware practices of big data location-based applications and services, providing a balance between privacy risks and data utilities.”

Among the study’s limitations, the authors note that the data they used contain no information about individual consumers’ demographics, which would allow greater understanding of privacy issues. In addition, their proposed framework considered only one-shot data sharing with an advertiser; it did not consider more complex scenarios with multiple risks or utilities, or what happens when an advertiser combines multiple batches or sources of shared data.