Confronting The Long-term Risks Of Artificial Intelligence
'The challenge lies in aligning AI with universally accepted human values'Photo Credit: Getty Images/iStockphoto
Risk is a dynamic and ever-evolving concept, susceptible to shifts in societal values, technological advancements, and scientific discoveries. For instance, before the digital age, sharing one's personal details openly was relatively risk-free. Yet, in the age of cyberattacks and data breaches, the same act is fraught with dangers. A vivid cinematic example of evolving perceptions of Artificial Intelligence (AI) risk is the film, Ex Machina.
In the story, an AI named Ava, initially viewed as a marvel of synthetic intelligence, reveals her potential to outwit and manipulate her human creators, culminating in unforeseen hazards. Such a tale exemplifies how our understanding of AI risk can drastically change as the technology's capabilities become clearer. This underscores the importance of identifying the short- and long-term risks. The immediate risks might be more tangible, such as ensuring that an AI system does not malfunction in its day-to-day tasks. Long-term risks might grapple with broader existential questions about AI's role in society and its implications for humanity. Addressing both types of risks requires a multifaceted approach, weighing current challenges against potential future ramifications.
Over the long termThe risks that present themselves over the long term are worth looking at.
Yuval Noah Harari has expressed concerns about the amalgamation of AI and biotechnology, highlighting the potential to fundamentally alter human existence by manipulating human emotions, thoughts, and desires. In a recent statement by the Center for AI Safety, more than 350 AI professionals have voiced their concerns over the potential risks posed by AI technology.
One should be a bit worried about the intermediate and existential risks of more evolved AI systems of the future — for instance, if essential infrastructure such as water and electricity increasingly rely on AI. Any malfunction or manipulation of such AI systems could disrupt these pivotal services, potentially hampering societal functions and public well-being.
Similarly, although seemingly improbable, a 'runaway AI' could cause more harm — such as the manipulation of crucial systems such as water distribution or the alteration of chemical balances in water supplies, which may cause catastrophic repercussions even if such probabilities appear distant. AI sceptics fear these potential existential risks, viewing it as more than just a tool — as a possible catalyst for dire outcomes, possibly leading to extinction.
The evolution to human-level AI that is capable of outperforming human cognitive tasks will mark a pivotal shift in these risks. Such AIs might undergo rapid self-improvement, culminating in a super-intelligence that far outpaces human intellect. The potential of this super-intelligence acting on misaligned, corrupted or malicious goals presents dire scenarios.
The challenge lies in aligning AI with universally accepted human values. The rapid pace of AI advancement, spurred by market pressures, often eclipses safety considerations, raising concerns about unchecked AI development.
The world does not have a unified approach. The lack of a unified global approach to AI regulation can be detrimental to the foundational objective of AI governance — to ensure the long-term safety and ethical deployment of AI technologies. The AI Index from Stanford University reveals that legislative bodies in 127 countries passed 37 laws that included the words "artificial intelligence".
One of the most celebrated regulations out of these is the European Union's AI Act. It adopts a 'risk-based' approach, tying the severity of risk to the area of AI deployment. This makes sense when considering AI applications in critical infrastructures, which demand heightened scrutiny. However, tying risk solely to the deployment area is an oversimplified strategy. It might overlook certain risks that are not directly tied to the deployment area. Therefore, while the area-specific approach is valuable, a more holistic view of AI risks is necessary to ensure comprehensive and effective regulation and oversight.
However, there is a conspicuous absence of collaboration and cohesive action at the international level, and so long-term risks associated with AI cannot be mitigated. If a country such as China does not enact regulations on AI while others do, it would likely gain a competitive edge in terms of AI advancements and deployments. This unregulated progress can lead to the development of AI systems that may be misaligned with global ethical standards, creating a risk of unforeseen and potentially irreversible consequences. This could result in destabilisation and conflict, undermining international peace and security.
Thus, nations engaging in rigorous AI safety protocols may be at a disadvantage, encouraging a race to the bottom where safety and ethical considerations are neglected in favour of rapid development and deployment. This uneven playing field can inadvertently encourage other nations to loosen their regulatory frameworks to maintain competitiveness, thereby further compromising global AI safety.
The dangers of military AIFurthermore, the confluence of technology with warfare amplifies long-term risks. Addressing the perils of military AI is crucial. The international community has formed treaties such as the Treaty on the Non-Proliferation of Nuclear Weapons to manage such potent technologies, demonstrating that establishing global norms for AI in warfare is a pressing but attainable goal. Treaties such as the Chemical Weapons Convention are further examples of international accord in restricting hazardous technologies. Nations must delineate where AI deployment is unacceptable and enforce clear norms for its role in warfare. In this ever-evolving landscape of AI risks, the world must remember that our choices today will shape the world we inherit tomorrow.
Aditya Sinha is Officer on Special Duty, Research, Economic Advisory Council to the Prime Minister. Tweets@adityasinha004. The views expressed are personal
Will Artificial Intelligence Replace Architects?
Will Artificial Intelligence Replace Architects?
Contemporary house in the Netherlands according to IA. Image, © Ulises Design Studio (@ulises.Ai) Share Facebook Twitter Mail Pinterest Whatsapp Or https://www.Archdaily.Com/1007802/will-artificial-intelligence-replace-architects "COPY" CopyWill Artificial Intelligence replace architects in their roles? In the May 2023 edition of Building magazine, Thomas Lane suggests that AI has the potential to automate up to 37% of the tasks typically carried out by architects and engineers. This automation, though, is likely to target routine and less creative tasks, allowing professionals to concentrate on more strategic and imaginative aspects of their work.
Just as Revit and 3D software did not replace architects but only transformed their workflows, the same principle holds for AI tools. AI is poised to bring about new tasks, such as AI management, alongside existing responsibilities, signaling a shift in how architects work.
In early 2023, the abundance of images generated by Midjourney and similar AI systems has left many architects pondering its implications. While there's a general concern about the possibility of Artificial Intelligence becoming all-powerful, architects are also intrigued and actively exploring how AI can be integrated into their practice, seeking to grasp its potential applications in their field.
It is improbable that AI will fully replace architects shortly. The architectural landscape is evolving rapidly, and although new applications will continue to surface, we are gradually gaining a clearer comprehension of AI's capabilities and boundaries. This developing understanding is shaping a more defined perspective of how AI can impact and transform our professional endeavors.
We have nothing to worry about until AI wins an architectural design competition.
Possible Future UsesHere are some broad categories that illustrate how Artificial Intelligence can complement architects and enhance our work:
1) Design Options Based on Specific Criteria
AI algorithms are highly skilled at handling vast amounts of data and, when equipped with the right tools, have the potential to generate design choices. In the future, architects will likely input criteria like budget, space needs, or sustainability objectives, enabling AI to produce optimized design alternatives for human review.
Casa contemporânea na Indonésia segundo IA. Imagem © Ulises Design Studio (@ulises.Ai)2) Site Analysis and Mass Studies
AI tools, such as Autodesk Forma, can extract urban information from open-access databases and help you quickly create mass studies considering environmental data, floor areas, building shapes, heights, setbacks, building codes, etc.
3) Generative Design:
AI can rapidly produce visual designs when given specific guidelines, including photorealistic images. Architects can leverage generative design software like Stable Diffusion, Midjourney, Dall-e 2, or Adobe Firefly to investigate different material choices and spatial designs during the initial project phases. Furthermore, hand-drawn sketches or existing images can be used as a starting point for creating more elaborate designs, iterating on them, or making modifications.
4) Pattern Recognition
AI can analyze and identify patterns within extensive datasets, spanning architectural drawings, technical configurations, and historical projects. Currently, the process of creating a database that's easily accessible and analyzable by AI is complex, but it is anticipated to become more manageable in the future as technology advances.
5) Coding - Custom Apps, Programs, and Plugins
AI can assist in coding and developing your applications and software, potentially aiding in organizing your design resources for building databases. It would be particularly beneficial if there were open-source code libraries where architects could collaborate and share information, as opposed to the current situation where libraries are often isolated and fragmented under corporate ownership.
6) Energy Efficiency and Sustainability
AI can help architects design energy-efficient and sustainable buildings. By analyzing factors such as climate data, building orientation, material, and energy consumption patterns, AI algorithms can recommend design modifications that make building energy use more efficient and reduce its environmental impact.
Contemporary house in Mexico according to IA. Image © Ulises Design Studio (@ulises.Ai)7) Data Summarization
AI can additionally assist in condensing information from books and reports, making it easier for you to locate essential details swiftly while still having the source available for review. ChatGPT is currently introducing plugins, initially accessible to Premium subscribers, that broaden the range of data that can be analyzed, including real-time internet data.
8) Building Maintenance
AI tools have been developed to analyze video feeds and detect weaknesses in areas such as road conditions and the exteriors of existing buildings, which enables early detection of required maintenance for infrastructure.
9) Bim and Project Management
Building Information Modeling (BIM) is a digital representation of a construction project, encompassing geometric data, materials, and other pertinent information. AI can analyze BIM data, spot potential conflicts or clashes, optimize schedules, and aid in project management activities, thus enhancing coordination and diminishing errors in the construction process.
10) Virtual Reality and Augmented Reality
Architects can harness the power of AI-driven virtual reality (VR) and augmented reality (AR) technologies to create visualizations and presentations of their designs. These immersive experiences enable stakeholders to navigate 3D virtual environments at various scales, even life-size (1:1 scale), to study spatial connections and make better-informed decisions regarding the building's design and layout.
11) Cost Estimation and Material Selection
AI algorithms can scrutinize historical cost data, building material specifications, and market trends, offering precise cost estimates and material recommendations. Architects can utilize this data to make well-informed choices that align with budget restrictions and project needs. To employ this, you might need to establish a database and establish a method for AI to access it unless an open-source or subscription-based database for such information becomes accessible in the future.
Contemporary house in China according to IA. Image © Ulises Design Studio (@ulises.Ai) Considering TrainingWhile AI is unlikely to cause mass job losses for architects shortly, it will fundamentally change the nature of our profession. To effectively embrace new tools and evolving knowledge, architects must remain current and continually update their skill sets.
The present moment calls for proactive discussions and solutions. Architects must begin leveraging AI to maintain a competitive edge in an ever more competitive environment. We must acquaint ourselves with AI's capabilities and elevate our proficiency to become "superusers."
When examining architectural education, we find numerous disciplines that delve into the history of architecture but few that contemplate its future. What architects ought to be conceiving today is not just another structure but the future trajectory of our profession.
Via Tabulla.
Can Artificial Intelligence Transform Public Health Nutrition?
In a recent editorial published in the Nutrients Journal, researchers described the applications of artificial intelligence (AI) to population well-being and nutrition.
Study: Artificial Intelligence Applications to Public Health Nutrition. Image Credit: metamorworks/Shutterstock.Com
BackgroundPublic health research covers a wide range of topics, from studying the effects of government measures such as taxing soda and nutritious food subsidies to examining how climate change and economic conditions affect food choices and accessibility.
In a well-connected world, fresh approaches and technologies are being developed to solve the issue of correctly monitoring nutritional intake on a broad scale.
Nutrition research has increasingly relied on artificial intelligence and machine learning models to comprehend, diagnose, forecast, and explain data. The capabilities and use of artificial intelligence have increased at an unparalleled rate.
Machine learning-based algorithms and exponential increases in processing capacity have fueled rapid progress and infiltrated and transformed a wide range of human endeavors and societal sectors. Based on training data, machine learning can enable a computer system to create an algorithm to translate input information into a specific output.
About the editorialIn the present editorial, researchers present AI's potential in public well-being and nutrition.
Use of artificial intelligence for public welfare and nutritionPublic welfare and nutrition take the lead in the overall fields of wellness and disease prevention, distinguishing itself from individual-level nutritional research.
The latter dives deeply into the intricacies of individual-level nutritional requirements, metabolism, and inheritance. In contrast, public welfare and nutrition are concerned with understanding and influencing the eating habits of whole populations rather than individual dietary needs and genetic predispositions.
AI's promise in public welfare and nutrition is being fulfilled in various novel ways. Artificial intelligence models have proved useful in visualizing and evaluating food surroundings and detecting places with restricted availability of nutritional foods, which are commonly referred to as "food deserts."
Machine learning algorithms are additionally used to forecast the consequences of future policy measures, such as the impact of certain subsidies or taxation on population food patterns.
On a larger level, artificial intelligence techniques have aided in monitoring worldwide food networks, ensuring food security, and predicting future disruptions caused by climate change.
AI enables tools to generate and analyze massive amounts of data, including satellite images of agricultural areas and online debates about dietary patterns, to enhance public welfare and nutritional interventions.
Limitations of using artificial intelligenceThere are serious concerns about artificial intelligence technologies that should not be neglected. Data security is critical, and any artificial intelligence-based public health initiative must guarantee that community and individual rights are respected.
Another key issue is bias; the accuracy of algorithms used in machine learning tools depends on the training data, and any prejudice in the training dataset might result in biased outputs, exacerbating existing health inequities.
While artificial intelligence and machine learning models are strong modeling tools, failing to apply thorough and well-thought-out modeling procedures can lead to erroneous findings and ethical and bias problems.
The community must capitalize on AI's benefits and devise novel methods to reduce its possibly negative consequences. It is a time of partnership between innovation and healthcare, and by collaborating, it can be guaranteed that artificial intelligence is a tool for positive change in improving nutritional health worldwide.
ConclusionsOverall, the editorial highlighted the application of artificial intelligence in the fields of public welfare and nutrition. The potential uses of artificial intelligence in population health and nutrition are numerous, and present research may just scrape the outer limits of what might be feasible.
In public welfare and nutrition, artificial intelligence is used to map and evaluate food environments, identify locations with restricted access to healthy foods, and anticipate the impact of policy actions on population eating patterns.
The abundance of data, along with AI's expanding capabilities, provides a plethora of unexplored opportunities. Researchers must think outside the box to leverage the potential of artificial intelligence to encourage better diets and enhance the nutritional state at the community level.
In the coming years, AI-powered models might anticipate the nutritional demands of entire areas according to soil quality, climate, and socioeconomic factors, supporting policymakers in determining food priorities.
AI tools could be used to track worldwide food patterns, detect novel trends, and assist public health professionals in drafting solutions faster. AI systems could even collaborate across nations to standardize nutritional requirements so that an identical message reaches everyone.
