New Report on Progress and Risks of Artificial Intelligence – OpenGov Asia

A new report showed that Artificial Intelligence (AI) has reached a critical turning point in its evolution. Substantial advances in language processing, computer vision and pattern recognition mean that AI is touching people’s lives daily—from helping people to choose a movie to aid in medical diagnoses.
With that success, however, comes a renewed urgency to understand and mitigate the risks and downsides of AI-driven systems, such as algorithmic discrimination or the use of AI for deliberate deception. Computer scientists must work with experts in the social sciences and law to assure that the pitfalls of AI are minimised.
The report – Gathering Strength, Gathering Storms: The One Hundred Year Study on Artificial Intelligence (AI100) 2021 Study Panel Report – aims to monitor the progress of AI and guide its future development. This new report, the second to be released by the AI100 project, assesses developments in AI between 2016 and 2021.
In the past five years, AI has leapt from something that mostly happens in research labs or other highly controlled settings to something that is out in society affecting people’s lives. The field is coming to grips with the societal impact of this technology. The next frontier is thinking about ways we can get the benefits from AI while minimising the risks.
– Professor of Computer Science, Brown University
Professor of computer science at the University of Texas said that what makes this report unique is the report is written by AI insiders – experts who create AI algorithms or study their influence on society as their main professional activity—and that they are part of an ongoing, longitudinal, century-long study. It also provides a wonderful template for future study panels to emulate by answering a set of questions that we expect future study panels to reevaluate at five-year intervals.
In terms of AI advances, the panel noted substantial progress across subfields of AI, including speech and language processing, computer vision and other areas. Much of this progress has been driven by advances in machine learning techniques, particularly deep learning systems, which have made the leap in recent years from the academic setting to everyday applications.
In the area of natural language processing, for example, AI-driven systems are now able to not only recognise words but understand how they’re used grammatically and how meanings can change in different contexts. That has enabled better web search, predictive text apps, chatbots and more. Some of these systems are now capable of producing original text that is difficult to distinguish from human-produced text.
Elsewhere, AI systems are diagnosing cancers and other conditions with accuracy that rivals trained pathologists. Research techniques using AI have produced new insights into the human genome and have sped the discovery of new pharmaceuticals. And while the long-promised self-driving cars are not yet in widespread use, AI-based driver-assist systems like lane-departure warnings and adaptive cruise control are standard equipment on most new cars.
Some recent AI progress may be overlooked by observers outside the field, but reflect dramatic strides in the underlying AI technologies, Littman says. One relatable example is the use of background images in video conferences, which became a ubiquitous part of many people’s work-from-home lives during the COVID-19 pandemic.
As reported by OpenGov Asia, in the healthcare field, U.S. Scientists have developed a new, automated, AI-based algorithm that can learn to read patient data from Electronic Health Records (EHR). The scientists, in a side-by-side comparison, showed that their method accurately identified patients with certain diseases as well as the traditional, “gold-standard” method, which requires much more manual labour to develop and perform.
The trial Mobile Money service approved by the Prime Minister will set a precedent for applying a “sandbox” scheme for new services and professions in the digital society. Sandbox is a controlled institutional framework applied to new technologies, products, services, and business models. It is an environment for technology firms to try their new technological apps and business models. After the trial period, management agencies will review the trial implementation and then accept or reject it.
Using laws to set rules to deal with new issues arising from the application of new technologies is a challenge. As per a press release, the apps may have a rapid impact on society that management systems may not be able to keep up with. Many traditional business fields have changed, and businesses have to utilise technology to work more effectively. It is impossible to manage new services and business models within the existing framework because policies tend to lag behind practices. Therefore, a sandbox model is more advantageous.
According to an industry expert, it is impossible to demand state management agencies to create policies for the future. Many countries apply sandbox policies to encourage enterprises to develop new business models, with certain limitations in deployment. The Prime Minister has put into effect the pilot implementation of Mobile Money services – making payments for small-value goods and services with telecom accounts. The pilot programme will last two years.
This is the first service that the government has applied the sandbox mechanism managed by several ministries and branches. The government hopes the service will contribute to the development of non-cash payments, and promote the access and use of financial services, especially in rural areas. Businesses can only provide Mobile Money to remit money and make payments for legal goods and services in Vietnam in accordance with current laws. Mobile Money is only applied to domestic transactions with a monthly transaction value limit of VND10 million (US$4,397).
Vietnam is not the first country that has accepted a new technology platform, but experts said that it has an advantage by learning lessons from predecessors. In Vietnam, the proportion of credit card users is still low, but mobile subscriber density is very high. 99% of transactions with a small value of below VND100,000 (US$4) are carried out in cash. Mobile Money will be a strong solution to promote non-cash payments in society.
The Minister of Information and Communications stated that Mobile Money is a convincing example that shows that telecom carriers can become platforms for many things, not only telecom infrastructure. They can become platforms for data, computing, digital content, authentication, IT services, and the Internet of Things (IoT).
Mobile Money is expected to help Vietnam become a digital society. The project is the first sandbox involving many ministries and sectors to be piloted to meet the needs of society. It will pave the way for more sandboxes to be applied to other new services and business models in the future. He added that Mobile Money is a great opportunity for mobile network operators to build an ecosystem to accelerate digital transformation.
The National Economic and Development Authority (NEDA) said it is accelerating the implementation of the Philippine Identification System (PhilSys) or the national ID programme to enhance the government’s ability to deliver various social services. NEDA chair said that more than 42 million Filipinos had registered as of September for step 1 or collection of demographic data. Notwithstanding the quarantines, he said nearly 30 million took the second step to supply their biometrics in the registration centres.
OpenGov Asia reported that the National Economic and Development Authority (NEDA) expects the Philippine Statistics Authority (PSA) to register 50 to 70 million people for the national digital ID by the end of the year. It is noted that as of July 2, 37.2 million people had completed Step 1, which involves the collection of demographic information, and 16.2 million had completed Step 2, which involves the capture of biometrics at designated registration centres.
The impact of the Covid-19 pandemic is challenging, but the Philippines has a solid foundation to recover at the right time. Reforms such as Rice Tariffication Law and the National ID are helping us restore our development trajectory and enabling the people, especially the poor, to access affordable food and better social services. 
– NEDA Chairperson
Speaking on the progress being made by the digital ID project, Socioeconomic Planning Secretary at NEDA said: “The COVID-19 crisis underscores the need to provide unhampered access to banking and social services for all Filipinos, especially the poor. Therefore, the President gave the directive to accelerate the implementation of the Philippine Identification System or PhilSys to provide all Filipinos with a unique and digitalised ID.”
He underlined that the Filipinos, particularly the poor, would be able to open bank accounts where cash transfers can be received directly. “We aim to register at least 50 million Filipinos by the end of this year,” he said.
The pandemic gave new promptness and highlighted the primacy of financial integration into government crisis containment and rehabilitation efforts. It showed the vital role of financial inclusion in social welfare and protection, as the transaction accounts became a necessary means to receive government cash support from the poorest and most vulnerable in the country.
As per NEDA’s Chairperson, PhilSys would also facilitate financial inclusion by providing every Filipino with a valid proof of identity, which is required for low-income earners to open bank accounts, receive cash transfers, and access other financial services.
Meanwhile, the Philippines’ state-run bank said it has signed up 5.3 million unbanked PhilSys registrants for their own transaction accounts via account opening booths at select PhilSys co-location areas nationwide. The registrants have already used their prepaid cards for a total of P31.8 million in transactions.
The PhilSys registrants can use the Landbank prepaid cards to manage funds, withdraw cash, perform cashless transactions, shop and pay bills online, and receive government subsidies digitally. As per the president and CEO of the Philippines bank, bringing unbanked Filipinos into the financial mainstream lays the groundwork for inclusive growth, particularly as we accelerate initiatives toward economic recovery and sustained development. Access to formal banking services motivates people to save money, repay loans, invest in financial products, and achieve financial independence.
Unbanked PhilSys registrants may access bank transaction accounts after completing the PhilSys Step 2 registration process, which includes validating supporting documents and capturing biometrics data. PhilSys registrants can also activate their bank prepaid cards and conduct transactions through the bank’s mobile branches, which are located in communities across the country where banking services are disrupted or limited.
The bank’s mobile branch is intended to serve unbanked and underserved communities as well as areas affected by disasters, calamities, and other disruptive events, as part of the bank’s increased efforts to promote greater financial inclusion in the community.
A research team led by biomedical engineers at the City University of Hong Kong (CityU) has developed a new generation of microneedle patches made of ice that melt after the pain-free delivery of drugs.
Experiments using this ground-breaking invention on mice with cancers have shown that the animals’ immune responses were much better than those seen in conventional vaccination methods. The technology paves the way for developing an easy-to-use cell therapy and other therapeutics against cancers and other diseases.
Made from a cryogenic solution, these icy microneedles are less than 1mm long and can deliver living mammalian cells into the skin. The device is like a skin patch and the microneedles can detach from the patch base, melt and then penetrate the skin.
The research is led by Dr Xu Chenjie, Associate Professor in the Department of Biomedical Engineering (BME), and the findings were published in Nature Biomedical Engineering under the title “Cryomicroneedles for Transdermal Cell Delivery”.
Dr Xu explained that traditional cell therapy for skin disorders is invasive, painful, complicated, low-efficient, risks infection, and requires experienced professionals. The ready-to-use device can circumvent complex and redundant procedures during each drug administration. In addition, it can be stored for months in a refrigerator and is easily transported and deployed.
The applications for this device are not limited to the delivery of cells. It can package, store, and deliver any type of bioactive therapeutic agents such as proteins, peptides, mRNA, DNA, bacterial, and vaccines, and it can improve both the therapeutic efficacy and patient compliance during cell therapies.
As a proof-of-concept, the researchers explored cell-based cancer immunotherapy through the intradermal delivery of ovalbumin-pulsed dendritic cells. Experiments showed that vaccination using therapeutic cells through this technology elicited robust antigen-specific immune responses and provided strong protection against tumours in mice.
These results were superior to the therapeutic outcomes of conventional vaccination methods. One of the start-up teams supported by the Seed Fund of HK Tech 300, CityU’s flagship innovation and entrepreneurship programme, is working on transferring the technology into a product and to promote its application.
Dr Chang Hao, a former postdoc in CityU’s BME, is the first author of this study, and Dr Xu is the corresponding author. Other researchers include Professor Wang Dongan and Professor Shi Peng from BME. The research team collaborated with scientists from Nanyang Technological University and the National University of Singapore.
The cell therapy technologies market is projected to reach US$5.6 billion by 2025 from US$2.8 billion in 2020, at a CAGR of 14.4% from 2020 to 2025. The emerging economies such as Australia and China are expected to provide a wide range of growth opportunities for players in the market which is driven by their large and growing populations as well as an increase in the number of clinical trials and investments in the field of personalized medicine in these countries.
The outbreak of COVID-19 is expected to have a minimal or negligible negative impact on the cell therapy technologies market. The rise in the incidences of COVID has led to an increase in the need for an efficient drug or vaccine for COVID, which could help in reducing the severity of the cases.
Cell-based research is an essential step during the manufacturing of vaccines, which can help in the growth of the market.
In the initial months of the outbreak of COVID, disruption in the supply chain had been witnessed, which has delayed the clinical trials. This can negatively impact the market to a certain extent. For instance, biopharmaceutical companies and major players have announced clinical trial delays.
A film is not complete without relevant and good music in the background. Music establishes atmosphere and mood and influences the audience’s emotional reactions as well as their interpretation of the story. A research team at the USC Viterbi School of Engineering sought to objectively examine the effect of music on cinematic genres. Their study aimed to determine if AI-based technology could predict the genre of a film based on the soundtrack alone.
While past work qualitatively indicates that different film genres have their own sets of musical conventions—conventions that make that romance film sound different from that horror movie—Narayanan and team set out to find quantitative evidence that elements of a film’s soundtrack could be used to characterise the film’s genre.
The study was the first to apply deep learning models to the music used in a film to see if a computer could predict the genre of a film based on the soundtrack alone. They found that these models were able to accurately classify a film’s genre using machine learning, supporting the notion that musical features can be powerful indicators in how people perceive different films.
This work could have valuable applications for media companies and creators in understanding how music can enhance other forms of media. It could give production companies and music supervisors a better understanding of how to create and place music in television, movies, advertisements, and documentaries in order to elicit certain emotions in viewers.
In their study, the team examined a dataset of 110 popular films released between 2014 and 2019. They used genre classification listed on the online database of information related to films to label each film as action, comedy, drama, horror, romance, or science-fiction, with many of the films spanning more than one of these genres.
They then applied a deep learning network that extracted the auditory information, like timbre, harmony, melody, rhythm, and tone from the music and score of each film. This network used machine learning to analyse these musical features and proved capable of accurately classifying the genre of each film based on these features alone.
The team also interpreted these models to determine which musical features were most indicative of differences between genres. The models didn’t give specifics as to which types of notes or instruments were associated with each genre, but they were able to establish that tonal and timbral features were most important in predicting the film’s genre.
The researchers examined the auditory information from each film using a technology known as audio fingerprinting. This technology allowed them to look at where the musical cues happen in a film and for how long. Using audio fingerprinting to listen to all of the audio from the film allowed them to overcome a limitation of previous film music studies, which usually just looked at the film’s entire soundtrack album without knowing if or when songs from the album appear in the film.
In the future, the team is interested in taking advantage of this capability to study how music is used in specific moments in a film and how musical cues dictate how the narrative of the film evolves over its course.
AI has been adopted in various areas, including healthcare. As reported by OpenGov Asia, U.S. Scientists have developed a new, automated, AI-based algorithm that can learn to read patient data from Electronic Health Records (EHR). The scientists, in a side-by-side comparison, showed that their method accurately identified patients with certain diseases as well as the traditional, “gold-standard” method, which requires much more manual labour to develop and perform.
Researchers from the Indian Institute of Technology in Bombay (IIT-Bombay) have developed a new data-processing technique to measure low amounts of soot accurately. This will help designers build better combustion-based devices such as internal combustion engines in cars.
Soot is tiny black particles that rise from a flame. Soot is formed when the fuel does not burn entirely. When fuel burns properly, a blue flame is emitted, whereas the flame is yellow when the soot is formed during burning and it becomes hot. Soot can cause cancer and respiratory and cardiac disorders and can also reduce the life of machine parts, a news report has explained.
Accurately measuring small amounts of soot can be a challenge and has spawned several research projects. The team from IIT-Bombay demonstrated a new technique to effectively reduce measurement errors when soot is present in low amounts. They analysed digital camera pictures of burning fuel to guess the temperature of the fuel and use the information to estimate the soot volume. The amount of soot can be measured using methods such as collecting and weighing the soot and studying a light beam shone on soot particles. The current study uses the last method. The researchers passed a beam of red laser light of a specific frequency, through a droplet of burning fuel and took images as it burnt. The light falling on the camera also contains the light from the burning fuel. The researchers used a narrow band filter to let only the laser light pass and filter out the light emitted by the burning fuel.
The report noted that when a flame having soot particles is shone with light, called background light, the particles absorb and scatter some of this light, so light reaching the camera is less bright. The researchers used the relation between the initial brightness of the laser light, the brightness of the light falling on the camera, and the soot volume to calculate the amount of soot. They then used a data-processing technique to compute the values of brightness from their images. Their challenge was to estimate the initial brightness of background light falling on soot particles since this isn’t directly captured in the images.
The team predicted the brightness of background light at every moment instead of using an average. They observed the flickers in background light at areas present outside the flame of the burning fuel, where there is no soot. They used it to estimate the background light falling on the soot particles. Using the new data processing technique, the team got lower errors, especially when the amount of soot produced is low. The technique does not require any additional equipment or extra expenditure, an added advantage.
The report added that to further reduce errors in the experiment, the researchers passed the laser light beam through a fixed and a rotating diffuser — a glass sheet that scatters light — before the light was incident on the burning fuel. A diffuser gives an evenly bright light and avoids the many speckles in the camera image. Speckles need to be removed while processing the data, leading to a loss of information. The researchers also validated their data processing technique. They used it to calculate the amount of soot for some previous measurements reported in the literature and verified the results. They also qualitatively checked their experimental observations.
They burnt a droplet of toluene (a carbon-based fuel) and compared their experimental observations with that in the literature. The team observed a similar peak value of the amount of soot. As expected, they saw high amounts of soot slightly inside the outer edges of the flame, where temperatures and fuel concentration are high, a researcher explained. The quantification of soot is crucial from an environmental perspective. This is an effective method to quantify soot to help identify strategies to mitigate combustion-based practices in India.
While the COVID-19 pandemic disrupted the economy, tremendous opportunities have also emerged in the post-pandemic economy to fuel the growth of startups in Taiwan. The outbreak of the pandemic in 2020 changed lives around the world, accelerating digital transformation and creating opportunities for new startups to develop innovative ideas and technological breakthroughs.
The Industrial Technology Research Institute (ITRI), which has support from the Ministry of Economic Affairs (MOEA), collaborated with the Stanford Centre for Professional Development (SCPD) to share their expertise and perspectives on technology innovation. The platform provided an opportunity for startups in Taiwan to learn about Silicon Valley’s entrepreneurial ecosystem and to reimagine business opportunities in the context of a ‘new normal’ economy.
The pandemic has impacted the global economic structure significantly, bringing about further developments in business innovation and digital technology. For the past few years, the Department of Industrial Technology (DoIT) has played an important role in promoting R&D innovation to help develop an entrepreneurial ecosystem in Taiwan
– Director General of DoIT, MOEA
The pandemic has escalated the importance of technology R&D. According to United Nations Technology and Innovation Report 2021, it forecasted that five leading technologies by 2025 would be the Internet of Things (IoT), robots, green energy, 5G and ArtificiaI Intelligence (AI).
These technologies may help mitigate some of the negative impacts of the pandemic while increasing productivity and improving life quality. They also represent a tremendous economic gain in the global market, allowing nations to transfer R&D capability into business opportunities. Through educational exchanges like this forum, Taiwan will learn how to cultivate an entrepreneurship culture and create new business opportunities for the global market.
During and after the COVID-19 pandemic, there will be major challenges for the medical care system, but this will also provide great opportunities for remote care services. Co-founder of health tech startup shared his entrepreneurial experience in the perspectives of ideation, needs finding, market planning, fundraising, and team composition, bringing inspiration to the technology R&D team. The startup is creating a non-invasive in-home device to facilitate screening and treatment for diabetic foot complications.
ITRI has been collaborating with various schools and centres at Stanford since 2004, and in 2014 they began working with SCPD to co-host these educational exchanges, focusing on sharing educational research and development while encouraging academic collaboration between Taiwanese and American educators. Achievements of such exchanges include facilitating Taiwan researchers to conduct research at Stanford as visiting scholars, co-authoring books between ITRI and international experts, and raising the international visibility of the successful experiences of Taiwan’s high-tech industry.
As reported by OpenGov Asia, Taiwan’s Minister of Science and Technology said that the global pandemic is accelerating digital transformation. The pandemic poses both challenges and opportunities to rethink technology development and supply chain management. Thanks to public and private efforts over the past year, Taiwan has largely withstood the ravages of COVID-19.
Infections spiked in mid-May, prompting various government restrictions. The number of COVID-19 cases is now declining. Taiwan has earned global praise for its effective campaign to limit the coronavirus, a program that includes an electronic fence that tracks residents via mobile phones to ensure that quarantined people remain at home. The effort is also seen as less intrusive than other tracking efforts.
Taiwan has shown its skills and expertise in public health and digital technology. With the belief that ‘Taiwan can help’, the government has been sharing resources and pandemic experiences with international partners. Many foreigners get to know about Taiwan in the process. The pandemic poses a challenge, but also creates an opportunity for Taiwan to highlight outstanding tech professionals and industry chains.
NSW Health-run vaccination centres are migrating to a new US-based company’s vaccination management system. eHealth NSW revealed the two-year contract with the American tech company for its Vaccination Administration Management (VAM) platform last week. The aim is to progressively migrate hubs off the in-house CoVax system.
VAM, which first went live in June, is replacing CoVax, a stopgap solution developed in-house over 21 business days earlier this year with help from other major global tech players. CoVax went live in February at South Western Sydney Local Health District (LHD) to support the frontline worker vaccinations and, by August, had expanded to all but three LHDs and speciality networks. VAM was configured and implemented in just 10 business days in anticipation of the accelerated vaccine rollout in the second half of 2021.
An eHealth NSW spokesperson stated that the tech company’s platform “ensures scalability and ease of use and is designed to manage the entire vaccination administration”. It is used to book an appointment, track vaccination doses, record clinical information, manage clinic workflows and interface with the Australian Immunisation Register operated by Services Australia. NSW Health chief information officer Zoran Bolevich has also previously said that the system has “simplified and automated workflows between administrators, clinicians and patients”.
Vaccination centres that were previously using CoVax are now being progressively migrated to VAM, a complex process as updates continue to be made and usage remains high. While the rollout is continuing, eHealth NSW last month said “appointments have already successfully transitioned from several live clinics, and speed and efficiency of this process continue to improve”.
Earlier this year, Victoria entered a six-month contract of similar value with Microsoft for its Covid vaccine management platform. Tasmania also has rolled out a health management system to allow residents to book vaccination appointments online.
The NSW government has released data showing Covid vaccination rates by area – individuals can use this interactive map to check their postcode or suburb. They can also toggle between showing the percentage of people who have received at least one dose and the percentage of people who are fully vaccinated. To compare the rates for areas over time to see the course of the vaccine rollout in that postcode, users can use the date menu.
New South Wales is currently experiencing the worst outbreak since the start of the pandemic in terms of the number of cases, with the more infectious Delta variant proving hard to contain. In another article, the Dubbo MP has praised the “real team effort” across the Western NSW Local Health District with the vaccine rollout in the region speeding up. Dubbo’s vaccination rate is around 85% for first doses and between 40-50 per cent for second doses.
The region has been extremely well-supported by, not just the local health district, but by the ADF, who have been here now for a few weeks delivering some of these vaccines in areas that would’ve been difficult to get the uptake as quickly without those clinics.