AI for Social Good: A map of United Nations efforts on AI

Here is a map of current UN activities in AI that I compiled from published UN reports. Everyone in AI has a responsibility to understand the opportunities to positively impact the world. If we don’t build responsible technology and solutions for our future generations, we will leave the world in shambles. We need to collaborate among ourselves to democratize technology, make it simple and build tools for solving technology adoption in areas where it is needed.

As you are reading this, think about what kind of tools can we build that makes adoption easy. More importantly, how can we collaborate to make progress on our efforts? There is no reason we cannot make progress as a community towards these goals; technology is mature in several of these areas.

Map of AI efforts by United Nations — https://www.saranyan.com/blog-pics/2019/8/19/un-ai-efforts

1. UN Department of Economic and Social Affairs (UNDESA)

1.1. AI areas

1.1.1. AI in eGovernment

  • Administrative (Filling forms, answering questions, routing requests, translation, drafting documents) tasks. Allows time for citizen engagement efforts.

  • Opportunities — Not to exacerbate issues around privacy, service delivery and ethics. Improve resource allocation and learn from historical data, which tends to mostly structured.

1.2. HLPF (High level political forum on sustainable development)

1.3. STI forum (Science, Technology and Innovation forum)

  • Forum focusses on SDG goals to connect technologists, stakeholders and identify technology gaps.

1.4. UN big data and official statistics

  • Working group that investigates the challenges and opportunities in Big Data including issues around data privacy, ethics, availability, etc.

1.5. Big data for public good

1.5.1. ECOSOC partnership — The Economic and Social Council (ECOSOC) is the United Nations’ central platform for reflection, debate, and innovative thinking on sustainable development.

2. United Nations Office for Disaster Risk Reduction (UNISDR)

  • Scanning the frontier technology horizon, what technological innovations are effectively ensuring resilience and DRR today?

  • How are frontier technologies being applied in vulnerable country contexts to ensure resilience?

  • What are the key barriers to scaling these technologies, noting typical challenges from capacity development and digital skills, finance and property rights areas?

  • How can governments address these barriers and accelerate the use of these solutions for DRR and resilience?

3. United Nations Conference on Trade and Development (UNCTAD)

3.1. Sustainable development

3.1.1. Using AI to accelerate SDG goals. Refer to my previous article on AI for social good.

3.2. Frontier technologies

3.2.1. Policy areas. Refer to UNCTAD policy hub.

3.2.2. International Tech and Trade Initiative (ITTI)

  • Businesses to trade (B2T) and leveling the playing field for SME. Example — Banks can use AI to make loan decisions to SMEs to reduce processing time. SMEs can integrate AI into their business processes for audit capabilities to stay compliant. AI products/startups have to stay on top of various government and international regulations.

  • Countries to trade (C2T): Allow nations to expand their competitive advantages. Using data to promote a given sector of the national economy both nationally and internationally. AI can speed up the process of information gathering. AI machines can can analyze data across WTO (World Trade Organization) rules to understand the pros and cons of different strategies, even suggesting the steps to be taken.

  • Negotiator to trade (N2T): Structured access to cloud-based resources to make life easier for trade negotiators. AI Negotiation support systems are able to suggest win-win outcomes. This can be adapted for trade negotiations.

  • Multilateral to trade (M2T): Multilateral trade officers will be able to weigh pros and cons of alternative scenarios through predictive AI. Automated analysis of trade, market, campaign and negotiation data to understand the impact of tariffs, investments, etc on the global market.

  • AI can help infrastructure (energy, networks, etc) to be shared among countries, communities or individuals. New business models for infrastructure are possible.

3.3. Focus on bottom billion

3.3.1. Understand the opportunities and challenges because of Frontier tech

3.3.2. Understand the impact on economy and society

3.4. Raise awareness of AI

4. UN programme on HIV/AIDS (UNAIDS)

4.1. AI areas

4.1.1. Information Source — Marlo Chatbot.

4.1.2. Understanding population trends and prediction of population growth.

5. United Nations Development Programme (UNDP)

5.1. AI Areas

5.1.1. Automation of Rapid Integrated assessment (RIA) (Evaluates national development priorities aligning to SDG targets)

5.1.2. AI to help the work of experts in development policy

5.1.3. Predict proxy levels of poverty (Use covariates from call details records to determine poverty levels)

5.2. Portfolio of frontier technology experiments

5.2.1. Drones and ML for environmental protection

5.2.2. Disaster preparedness

5.2.3. Mapping of refugee settlement (Develop infrastructure for refugees)

5.2.4. MapX: Catalog of best available spatial data and tools to map and monitor sustainable use of natural resources.

  • Partnership between UNDP, World bank, UNEP, Global information DB

  • Cloud based geospatial solution

5.3. UN Biodiversity lab

5.3.1. Partnership between UNEP and UNDP (Spatial analysis platform to enhance decision making on conservation)

5.3.2. Goal — Accelerate biodiversity targets (Aichi biodiversity targets)

6. United Nations Economic Commission for Europe (UNECE)

6.1. AI areas

6.1.1. WP.29

  • Regulations for autonomous vehicles

  • Vehicle management

  • Limiting wrong use of AI

6.2. Future networked car event

6.2.1. UNECE and ITU partnership

6.2.2. Status and future of vehicle communications and automated driving

6.2.3. Ethical considerations in AI and autonomous systems

6.3. Urban KPIs for smart cities

6.3.1. Circular cities: Green infrastructure that ensures mitigation of waste. Green infrastructure is a strategically planned network of natural and semi-natural areas that feeds into a circular utilization of resources and reduction of waste. AI is useful for planning.

6.3.2. Financing smart sustainable cities

6.3.3. Blockchain and AI in cities

6.3.4. Sensing technologies and IoT in cities: Connected communities through smart sensors and decision making. Through AI, it is possible to understand how cities are being used.

7. United Nations Interregional Crime and Justice (UNICRI)

7.1. Crime prevention

7.2. AI areas

7.2.1. Threat detection and landscape: Smarter, autonomous security systems that learn without human intervention and keep pace with the amount of data the security systems produce.

7.2.2. Behavioral patterns of terrorist networks: Detect pattern anomalies. Monitor new technology landscape to ensure preparedness. Advance understanding of and prepare for the risk of malicious use of AI by criminal and terrorist groups.

7.2.3. Predictive policing

7.3. Trust in AI and robotics

7.3.1. Ethics and Legal issues: Law enforcement using AI should take steps to ensure fairness, accountability and transparency. The use should be communicated to communities.

7.3.2. Algorithmic bias

7.3.3. Explainability (non black-box)

7.4. Criminal justice

7.4.1. Public safety video and image analysis.

7.4.2 . Scene understanding: Ability to develop text that describes the relationships between objects in a series of images to provide context.

7.5. Law enforcement AI use cases

7.5.1. Autonomously research, analyze and respond to requests for international mutual legal assistance.

7.5.2. Advanced virtual autopsy systems

7.5.3. Autonomous robotic patrol systems

7.5.4. Predictive policing and crime hotspot analytics for optimizing law enforcement resources

7.5.5. Computer vision to identify stolen cars

7.5.6. Tools to identify vulnerable and exploited children

7.5.7. Behavior detection tools to identify shoplifters

7.5.8. Fully autonomous tools to identify and fine online scammers

7.5.9. Crypto based packet tracing tools enabling law enforcement to tackle security without invading privacy.

8. United Nations Environment Programme (UNEP)

8.1. AI areas

8.1.1. Fintech for sustainable development: Less vulnerable financial systems, Creation of new markets, minimizing the risks and maximizing opportunities because of robotic automation.

8.1.2. Animation of the physical world: Using IoT and AI to connect the physical and natural assets, machines and infrastructures together will allow them sensing and responding to each other.

8.1.3. Planetary data governance

  • World environment situation room, a new data platform, but also a new way of accessing data, via a worldwide partnership model with multiple data-centers

8.1.4. Planetary dashboard for Surface water monitoring.

  • Organizations partnering the efforts — NASA, ESA, Google Earth, JRC

9. United Nations Educational, Social and Cultural Orgnaization (UNESCO)

9.1. Ethics

9.1.1. Ethical norms and standards: Using ROAM frameworks (Rights, Openness, Accessibility, Multi-stakeholder governance) to evaluate AI.

9.2. Policy

9.2.1. Freedom of expression, privacy and inequality

9.2.2. Safe and beneficial use of AI

9.3. Capacity Building

9.3.1. Counter the knowledge divide and marginalization of people

9.3.2. Create awareness around AI science and technologies

9.4. Platform for ethical dimensions on AI

9.4.1. Reflect on how AI could transform societies

9.4.2. Risks and benefits of transformations

10. United Nations Population Fund (UNPF)

10.1. GRID (Geo-referenced Infrastructure and Demographic Data for Development)

10.1.1. Access to spatial datasets for evidence-based and humanitarian decision making

10.1.2. High resolution spatial reference data

  • Population

  • Settlements

  • Infrastructure

  • Boundaries

10.1.3. Spatial modeling to generate accurate high-res population maps

10.2. Partnerships

10.2.1. UNFPA

10.2.2. Bill and Melinda Gates Foundation

10.2.3. DFID

10.2.4. Flowminder/Worldpop

10.2.5. Oak ridge national lab

10.2.6. Center for international earth science network

11. Comprehensive Nuclear-Test-Ban Treaty (CTBT)

11.1. International Monitoring System (IMS)

11.1.1. 337 facilities worldwide (monitor the planet for signs of nuclear explosions)

11.2. International Data Centre (IDC)

11.2.1. acquires data from the IMS global monitoring stations

11.2.2. Data distribution

11.3. On-site inspections (OSI)

11.4. AI efforts

11.4.1. Seismic phase

  • Data processing on Seismic signals

  • Classifiers to determine nuclear activity

  • Improve performance of classifiers

  • DNN architectures

  • Different data like waveforms

11.4.2. Event detection

11.4.3. Satellite monitoring (Change detection in inspected areas)

11.4.4. Seismic aftershock monitoring (Changes in the geological structures caused by a possible nuclear explosion)

11.4.5. Operations for sustainment of IMS

12. United Nations Children’s Fund (UNICEF)

12.1. AI Areas

12.1.1. Magic box (Open source platform that combines new sources of data in computational modeling to generate insights like spread of epidemic)

12.1.2. Project connect (Satellite imagery and DL for infrastructure mapping)

  • 130000 Schools mapped in 9 countries

12.1.3. Deep empathy (Use AI to increase empathy for victims in far-away disasters)

12.1.4. Generating equitable data sets

  • Training equitable AI algorithms

  • Compilation of symbols from different languages and cultures for children with disabilities

12.1.5. Venture fund investments

  • Capacity building

12.1.6. Building internal knowledge and capacity

12.1.7. Drone imagery and AI (Improve response during outbreaks)

12.2. Policy change

12.2.1. Use AI and NLP to understand and analyze constitutions from 194 countries

  • Advocate for human and environmental rights

12.2.2. Research on impact of AI on economy

12.3. Future work

12.3.1. Chat bot, social messaging platforms

12.3.2. Analyze implications of child labor in supply chain

12.3.3. Optimizing transactions and communication flows in organizations (Train on internal datasets)

13. United Nations High Commissioner for Refugees (UNHCR)

13.1. Predictive analytics

13.1.1. Project Jetson (Predict population movements in the horn of Africa, ex. Somalia)

13.2. AI in HR

13.2.1. Screening candidates to the talent team

14. United Nations Global Pulse

14.1. Achieve critical mass of high potential applications in AI and big data

14.1.1. Refer to UN Global pulse project page

14.2. Lower systematic barriers to innovation

14.3. Strengthen the data innovation ecosystem

14.4. AI areas

14.4.1. Using speech recognition technology to inform on SDG related topics in Africa

  • Convert public radio discussions to local languages

14.4.2. AI to detect structures in Satellite images to mark humanitarian efforts

14.4.3. Haze Gazer — a crisis analysis tool

  • Use satellite images and population data to enhance disaster management efforts

  • DL to determine air quality by fusing meteorological data, satellite imagery and social media pics

15. International Civil Aviation Organization (ICAO)

15.1. AI areas

15.1.1. Analysis of aviation infrastructure in terms of readiness for responding to disasters

15.1.2. Predictive model for preventing the spread of communicable diseases through aviation

15.1.3. Natural Language interface for decision-makers to interact with safety information

15.1.4. Neural Network application to classify Notice to Airmen (NOTAM) to reduce the noise

16. International labour organization (ILO)

16.1. AI areas

16.1.1. Impact on jobs and inequality

  • Risks, trends, analysis

  • Governance and infrastructure gaps

  • Working conditions in micro task platforms

16.1.2. Big data analytics

  • Capabilities of middle-income and emerging economies

  • Opportunities for economic diversification and new technology adoption

  • Equal access to job opportunities

16.1.3. Skill development

  • AI tools to measure current and potential skill demand

  • Diagnotsic tools

  • Understanding evolution and composition of job tasks

  • Understanding skills and knowledge content of occupations

  • Job matching policies

16.1.4. AI to improve learning delivery

16.1.5. Framework for public employment creation policy

16.1.6. Monitor child labour

  • Combating child labour and human trafficking

  • assess situation of the child and provide recommendations

17. International Telecommunications Union (ITU)

17.1. AI for good global summit

17.1.1. Accelerate development of AI solutions

17.1.2. Democratize AI development

  • Empower people to address social problems

  • Poverty

  • Hunger

  • Health

17.2. ML infra for 5G

17.2.1. Functional network architectures: 5G networks are far more complex than previous generation networks. Higher frequency radio technology, complex antenna configurations, sophisticated connectivity mechanisms like beamforming, dynamic and elastic network resources, etc require machine learning approaches. AI will play a significant factor in the design, deployment and monitoring of 5G network infrastructure.

17.2.2. Interfaces: Data and connectivity, Reducing gaps between edge and cloud computing.

17.2.3. Protocols and algorithms: Self-healing and resilience, Algorithms for streaming quality improvements, Detection of leakage from HFC networks, etc.

17.3. AI for health

17.3.1. ITU

  • ITU focus group on AI health

  • Standardize evaluation and validation of ML algorithms

  • Develop benchmarks

  • Standards frameworks

  • AI for health use cases

  • Mobile diagnostics

  • 6B smartphone deployments by 2021

  • ITU briefings on AI

  • Publication on AI

  • Global AI repository

17.3.2. WHO efforts

  • Access to wellness

  • Tracking outbreak of infectious diseases

  • Modeling and treatment of chronic diseases

  • Collection and storage and sharing of large datasets

18. United Nations Institute for Disarmament Research (UNIDIR)

18.1. Impact of AI on international Security

18.2. Weaponization of AI risks

18.2.1. Functional concerns like accidents in deployment

18.2.2. Manipulation and weaponization

  • Vulnerabilities

  • Datasets

  • Algorithms

  • Algorithm-driven weaponization of information

  • Conflict and stability

18.2.3. Commercial development

18.3. Mitigating harm from AI weaponization/security

19. United Nations Industrial Development Organization (UNIDO)

19.1. SAP digital boardroom

19.1.1. Monitor progress on SDG 9 (Sustainable industrialization)

19.2. Knowledge sharing thorugh Global Forum events

19.3. Understand role of AI in convergence of technologies

19.4. Industrial 4.0 efforts, opportunities and impact of AI

19.4.1. Industry 4.0 center in South Africa

19.4.2. Realtime monitoring of energy efficiency

19.4.3. Framework for industrial parks, eco parks

19.4.4. Bridge for City

References

  1. UN efforts on AI.

  2. AI for citizen services and Government — Harvard

  3. Webinar on Government Innovation and Disaster Risk Reduction

  4. McKinsey notes on AI frontier

  5. New Frontier of competitiveness in emerging technologies

  6. ITTI Trade initiatives

  7. Risks and benefits of AI and Robotics — UNICRI report

  8. AI and Robotics for Law Enforcement — UNICRI report

  9. UNEP — Fintech and sustainable development

  10. UNESCO — Steering AI for knowledge societies

  11. UN Global Pulse projects

  12. Machine learning for 5G future

Democratizing AI

This article appeared in insidebigdata

Artificial Intelligence (AI) is a technology accelerator and enabler. When we truly democratize AI, we have the opportunity to cause a massive paradigm shift in how humanity solves problems.

In 2016, TensorFlow showed the power of democratizing AI when a cucumber farmer got excited by the potential of using deep learning for sorting cucumbers. Outside the fact that this put TensorFlow on the map, it showed us a glimpse of what the world would look like when AI is easy to understand, use and access.

(Full article here)

Limiting bias and inexperience in AI-powered factories of the future

Published in techtarget.

The United Nations Sustainable Development Goals eight and nine are important in the context of Industry 4.0 and industrial IoT. SDG-8 calls for decent work and economic growth, while SDG-9 calls for innovation in industry and infrastructure. The purpose of the SDGs is to improve social conditions and advance humanity. AI plays a critical role in accomplishing this. For instance, let’s look at the innovation that’s happening in the Industry 4.0 space and where AI systems are proving efficient in preventing human errors and improving efficiency. The case studies from early AI systems clearly demonstrate that AI can not only improve efficiency metrics, like yield and throughput, but it can also reduce material waste and harmful emissions. In these scenarios, AI will create a net gain for us as society, improving human conditions.

Full article here

Data is not equal to knowledge

Published in Manufacturing.net. Full article here

A common pitfall a lot of machine learning (ML) companies run into is mistaking data as knowledge. Several enterprises think that having a lot of data makes them ripe for harvesting insights instantly through AI and ML techniques. It is not entirely true.

Data is not equal to knowledge, or more precisely, not the knowledge you think it equals.

Ernesto Miguel, 47 is a plant operator in a leading cement company. He has spent the last three decades working in the same cement plant. He knows each and every machine in his cement plant intimately. From the sound they make, he can tell what can be wrong. He is a champion in ensuring that the machines operate at their highest efficiency.

Full article here

Limiting bias and inexperience in the AI-powered factories of the future

This article originally appeared on techtarget as an invited guest article.

The United Nations Sustainable Development Goals eight and nine are important in the context of Industry 4.0 and industrial IoT. SDG-8 calls for decent work and economic growth, while SDG-9 calls for innovation in industry and infrastructure. The purpose of the SDGs is to improve social conditions and advance humanity. AI plays a critical role in accomplishing this. For instance, let’s look at the innovation that’s happening in the Industry 4.0 space and where AI systems are proving efficient in preventing human errors and improving efficiency. The case studies from early AI systems clearly demonstrate that AI can not only improve efficiency metrics, like yield and throughput, but it can also reduce material waste and harmful emissions. In these scenarios, AI will create a net gain for us as society, improving human conditions.

AI can transform humanity by giving time back to humans to focus on more productive tasks. There are new skills to be learned and it is clear that specific types of work will be displaced by new ones. For the sake of this article, let’s assume that we are able to empower our current factory workers with new skills that make them relevant and productive in the age of AI. If we do that, are we all set? Is that the only societal challenge we have for realizing the potential of AI completely?

In an ideal world, there are AI systems working seamlessly with humans to create factories of future that are lean, efficient and environmentally friendly. But we are far from that ideal world for two reasons: the current infrastructure present in industrial setting to collect and provide accurate data, and algorithmic biases.

There are different ways of architecting AI systems. The most common way is to model the behavior of the world through data and make decisions based on the realized model of the world. As you can see, this is problematic. What if the data is not accurate? What if we don’t have enough data? What if our data only partially captures the world we want to model?

With the last surge of industrial IoT revolution, there was a surge of dataavailable in factories. This opens the door to applying AI to factory operations. The challenge, however, is that the data is not ideal in several ways. Data collection processes were never optimized for a future AI application, rather they were built for simple responsive actions and decision-making. This shows up when the data is used to create machine learning models for building smart automation or predictive maintenance tools. Some problems with data can include incorrect sample rate, compressed or lossy data, incorrect sensor readings through faulty sensors or mechanical degradation, and so forth.

Algorithmic bias in AI, simply put, is a phenomenon where an AI deployment has a systematic error causing it to draw improper conclusions. This systematic error can creep in either because the data used to model and train the AI system is faulty, or because the engineers who created the algorithms had an incomplete or biased understanding of the world.

There have been several articles published about the human bias contributing to biased AI systems. There is well-documented evidence of AI systems showing biases in terms of political preferences, racial profiling and gender discrimination. However, in the context of Industry 4.0 applications, they are as big of a problem as data bias.

Going back to the SDG goals discussed above, we should aspire to improve the human conditions by providing people meaningful work. Let’s take an example of Ernesto Miguel, who has worked at a cement factory as a plant operator for the last 30 years. Ernesto spends most of his time ensuring the equipment under his watch functions efficiently. Over the last three decades, he has formed an intimate bond with the machines in his factory. He developed extraordinary abilities to predict what might be wrong with a machine by hearing the sound it makes. He can do more, like training more workers to be intuitive like him. He wants to share his expertise, but unfortunately Ernesto spends most of his time reacting to equipment problems and preventing failures. This is a problem ripe for AI.

We deployed one of our AI systems to model a crucial piece of plant equipment — a cooler — in a cement factory. The idea was to learn how adequately we could model equipment behavior by looking at two years’ worth of time series data. The data provided a great deal of insight into how the cooler was operating. Using the data, our engineers were able to identify correlation between different inputs to the equipment and its corresponding operating conditions.

If this worked flawlessly, we would accomplish two goals: use smart AI systems that could keep the equipment functioning in an optimum way and allow Ernesto to focus on more meaningful work, such as effectively training other factory workers.

Bias creeps in inadvertently when AI system designers confuse data with knowledge.

It was a big moment when the first AI system was deployed in the cement plant. We don’t yet live in a world where we can trust machines completely, and for good reason. So, there was a safety switch included for the plant operator to intervene if something went wrong. The first exercise was to run the software overnight, where the AI system monitored the cooler and was responsible for keeping it within safe bounds. To the delight of everyone, the system successfully ran overnight. But that joy was short-lived when the first weaknesses in the model started appearing.

The cooler temperature was increasing. And the model with an established correlation between the temperature and fan speed kept increasing the fan speed. In the meantime, the back grate pressure rose above the safe value. But the model identified no correlation between the back grate pressure and the temperature and felt no need to adjust the back grate pressure in its objective of bringing down the cooler temperature. The plant operator overrode the control and shut off the AI model.

An experienced plant control would have immediately responded to the increasing back grate pressure as it is detrimental to the cooler’s operation. How did the AI model miss this?

In his 30 years, Ernesto never had to wait for the grate pressure to build up before reacting. He just knew when the pressure would build up and proactively controlled the parameters to ensure that the grate pressure would never cross a safe bound. By merely looking at the data, there was no way for the AI engineers to determine this. The data alone without context would tell you that the grate pressure would never be a problem.

Bias hurts AI systems in many ways. The biggest of all is that it takes trust away from these systems. On top of watching his workers and equipment, Ernesto will have to watch the AI models. He has to teach the system to do things differently, which the system then has to learn. The next versions will improve. This will always be a problem when we model AI systems purely from incomplete or inaccurate data. In industrial IoT settings, this will always be the case because data will be inaccurate or incomplete.

As technology builders, what does this mean for us? How do we realize the full potential of industrial AI systems? The answer lies in us starting to design these systems with empathy and taking a thoughtful approach:

  • We cannot assume that data is a complete representation of the environment we are aspiring to model.

  • We need to spend time doing contextual inquiry — a semi-structured interview guided by questions, observations and follow-up questions while people work in their own environments — to understand the life of the workers who we are trying to empower AI systems with.

  • We need to assess all the possible scenarios that could occur in the problem we are trying to solve.

  • We need to always start with a semi-autonomous system and only transition to fully autonomous system when we are confident of its performance in production environments.

  • We should continually adapt and train models to learn about the environment we are operating in.

Bringing AI into factory settings is more than just technology. It is about people. It is also about doing something with empathy and understanding the people whose lives the technology is going to touch.

On AI democratization

In June 1993, NCSA Mosaic was launched. It was one of the first graphical browsers that was instrumental in popularizing the world wide web. It had a clean user interface and ran on Windows. It brought the power of internet mainstream, and became truly a killer application.

Browsers democratized internet. What will democratize Artificial Intelligence (AI)?

Today, many companies are working on AI platforms. Several companies are claiming to (or wanting to) democratize AI. What does this mean?

The internet provides us infrastructure to create and consume information. By doing this, it lets us collaborate and forge stronger communities. The browsers made this easier and brought the value of internet to everyone with access to a computer. Democratization, in this context is access to internet. Another way to understand browsers is that they allowed internet to be used effectively.

Working on the same parallel, the first thing we need to understand is the purpose of AI. Why does AI exist and what does it enable in this world, which if left untapped, causes human potential to be unfulfilled?

A common understanding of the above question might be the clue to understanding how to democratize AI.

A text book definition of AI is along the lines of creating agents that achieve their objectives by performing a sequence of actions, or exploring a sequence of operations. Machine Learning, which is often confused with AI these days, is only an aspect of AI where data from real world is used to train an AI system on some truth. 

Machine learning (ML) has a clear purpose of advancing human decision making capabilities based on prior evidence or data. For this reason, ML platforms will continue to be successful. At some point not in distant future, we will see a platform that will truly make ML mainstream. It will be similar to what NCSA Mosaic did for world wide web. Some argue that the current ML platform tools and frameworks have already brought ML mainstream. I don’t think it is true. An ML platform that truly abstracts the technicalities and focuses on a core human purpose will help democratize ML. A platform that truly understands and improves human productivity might be the killer app for ML.

What is the purpose for AI? Along the same lines, we can safely assume that like with every technology, our intent is to advance the human race and elevate it to its full potential through AI. If ML gives us superhuman capabilities to observe the world and make decisions based on it, AI might leverage that learning to make decisions on our behalf. 

The last point captures both the promise and peril of AI. While the prospect of observing the world and taking actions (that fulfill our objective) is a thrilling idea, it puts onus on us to architect objectives that are aligned with our human values and potential. It requires us to choose well and be aware of the implications of our choices. 

Then, can an AI platform essentially be a value framework that ensures that we don’t mess up? Can it be something that reminds us to construct objectives that are aligned with human values? Can there be a browser equivalent for an AI platform that lets people consume, create and collaborate on shared objectives that makes us better human beings? 

However, before such a product manifests, several things need to happen. We will have to put some basic infrastructure in place to support the creation and growth of such AI systems in our society. Tactically, we might need to create easy ways to consume and contextualize any data we interact with. This will need standard interfaces. Essentially, we will have to develop some protocols and shared language around how we understand these systems. In the process, we will create and optimize a wide array of workflow tools that allows us to build ML algorithms without writing code. An interesting argument can be made here that, if we truly mature in creating such ML frameworks and allow machines to design the right workflow and/or algorithms for solving an objective in the presence of reasonable constraints, we might be talking about the beginnings of a true AI system. Such a system would be able to identify a problem, explore data relevant to that problem, train itself in decision making and make decisions.

That might be one path towards AGI.

Special thanks to Eric Xing and Devin Sandberg for reading this article and providing feedback.