Artificial Intelligence and MSD/VCD interventions

1.?? Background

Artificial Intelligence (AI) is the biggest buzz word at present in pretty much all industries. It is being considered as the most transformative after fire or wheels or steam engine or sliced bread or internet; depending on whom you are asking. There are also emerging groups who are raising strong and very valid questions about the actual potential of AI considering cost & benefits. I would like to clear my standpoint that I am not in either of the groups rather someone who is exploring AI with an open mind. The benefit I have is the years of experience at the cross section of Market Systems Development (MSD), Value Chain Development (VCD) and Digital Development/ICT4D. A technology with social science background and years of field experience surely helps.

I have been involved in implementing several MSD and VCD projects across different markets in Africa, Asia and the Pacific. These markets include palm oil, fisheries, fertilizer, vegetables, food processing, seed, SME financing, ICT market and public services.

In most of these markets there were elements of digital, in many cases digital was a crosscutting element and in many of the projects focused on bringing in digital solutions for bringing in systemic change. It may sound very novel but in practicality these are very complicated, difficult and with many failed attempts. The failure occurred for several reasons starting from lack of technological readiness, access to the technology, trust in the technology and unable to create a commercially sustainable model. ?

Let’s dig into MSD/VCD and artificial intelligence possibilities.

2.?? Definition of Market Systems Development (MSD) and Value Chain Development (VCD)

If you are familiar with or a practitioner of MSD/VCD please feel free to skip this section. For the readers who are not familiar with MSD & VCD, I will try to define these in brief so that when these are used later in the discussion, it would be easier to follow.

Market Systems Development (MSD) and Value Chain Development (VCD) are two crucial approaches in international development aimed at creating inclusive, sustainable, and competitive markets. MSD focuses on improving the overall market environment by looking at the market as a system, making markets work better for the poor by addressing systemic constraints and promoting sustainable changes. MSD is also known as Making Markets Work for the Poor (M4P). And MSD is considered the main approach that led to the development of Bottom of the Pyramid (BoP) approach. BoP is more targeted, focused at developing consumer markets for the marginal population with the aim of improved liveilihoor and market creation. ?MSD example could be agro-processing market or finished leather market or skill development market. One MSD can contain multiple value chains in the system. At the same time MSD can focus on a single product like rice, potato, pottery, distribution services etc.

VCD, on the other hand, emphasizes enhancing specific value chains, ensuring that all actors, especially smallholders and SMEs, benefit from improved market access and efficiency. Generally, VCD’s are considered from immediate pre-production or production stage to the final consumer. In VCD the focus is on a particular item ‘s value chain like rice, wheat, seed, oil, woodwork, light engineering, meat etc. If looked closely the similarities with MSD are clearly visible.

There are many similarities and there are differences between these two approaches. It can be said that VCD focuses more on one product and MSD views the whole market. And in MSD approach it often happens that to achieve a systemic change may be one component outside the direct value chain of that market needs to be addressed.

If you want you can read more on BeamExcahge or Enterprise Development or The Springfield Centre websites for resources and case studies.

3. Introduction to AI in Development

Artificial Intelligence (AI) is being the hottest technological development in 2024, all the industries are trying to figure out how to use it and international development is no different. The Oxford Leaner’s Dictionary defined AI as “the study and development of computer systems that can copy intelligent human behaviour”. This is surely a very lucrative potential for businesses where the workforce can be deployed 24/7. However, the data analysis capabilities of are the key benefits as AI can process a large amount of data to identify patterns or come up with insights that would be very difficult or impossible for humans to do. ?Along with the corporations the government and international development agencies are wondering how AI could be used to tackle some of the world’s most complicated social issues. Recently released documents by USAID, FCDO on digital development have put significant importance of AI enabled development in the near future. However, these are all still in theory and we are yet to see a strong impact of AI in the international development addressing some challenges that could not have been solved otherwise. There are some initiatives, and we are expecting to see some breakthrough successes.

4. Understanding the Potential of AI in MSD/VCD

AI can be vital for many industries, however for AI to be useful for the marginal population we need to understand how AI can make the markets efficient so that the market can be impacting on the lives of the marginal population in a positive manner.? A subsistence farmer in Nigeria or Bangladesh or Samoa does not need to use AI or become proficient in prompt engineering, rather it is the support services providers who can use AI for developing customized solutions that fit the farmer’s need. I feel the following would be the areas where AI can make a huge difference and the eventual impact would be on the target population. The areas are

??????? I.??????????? Data Analysis and Service Customization

????? II.??????????? Market Information Systems

??? III.??????????? Supply Chain Optimization

?? IV.??????????? Financial Inclusion

???? V.??????????? Regulatory Oversight

Let’s discuss these briefly.

?I.??????? Data Analysis and Service Customization

By processing large datasets and providing actionable insights, AI can address some of the key challenges in identifying patterns, predicting outcomes, identifying information asymmetry, inefficiencies and insights. This was not possible before due to the sheer data volume and the human engagement required before. At present, with the support of AI the data analysis has become much simpler and effective. The insights derived through AI supported data analysis would help market actors to develop solutions that can cater to different segments.

II.??????? Market Information Systems

AI can enhance market intelligence by collecting and analyzing real-time data. This is particularly beneficial for smallholder farmers and SMEs, who often lack access to timely and accurate market information. AI-driven platforms can provide insights into market prices, supply and demand dynamics, and competitive analysis, empowering these actors to negotiate better prices and make strategic decisions. I don’t see farmers getting immediate benefits but with the involvement and investment of the development sector systems could be developed that would benefit the farmers.

?III.?? Supply chain optimization

AI can optimize supply chains by improving logistics and inventory management. AI algorithms can predict demand, optimize routes, and manage inventory levels, reducing wastage and costs. For instance, AI-driven logistics platforms can streamline the transportation of goods, ensuring timely delivery and reducing spoilage, which is crucial in agricultural value chains. For the contract-farming scenarios this can help farmers to have better support from the contract-buyers. However, most beneficial would be the large corporations as they can have very precious economies of scale on different supply chains.

?VI.???? Financial Inclusion

This is one area where AI can have a significant impact on customizing solutions and offers for the individuals and micro enterprises. This is again due to the superior capabilities of data analysis of AI to identify dynamic credit ratings, approvals for micro loans or scam prevention with better security mechanisms. Specially in the case of mobile financial services AI can play a very important role to improve overall service, product customization and improved consumer protection.

V. Regulatory Oversight

At present in most of the developing countries regulators rely on manual data collection, processing and then decision making. Though there are some digitization efforts and success but leveraging the power of AI can help the regulators to investigate granular data, identify more patterns and improve data-based decision making to effective regulation to ensure the market benefits the marginal populations.

5. Challenges and Considerations

Though the implementation of AI sounds fascinating, there are significant challenges and multi-dimensional considerations that need to be considered before developing any AI solution.

I. Technical and Infrastructural Barriers: Data quality and availability are critical issues, as AI relies on accurate and comprehensive data. Additionally, infrastructure limitations in rural and remote areas can hinder the effective deployment of AI technologies. Without proper data collection mechanisms and ensuring continual availability of quality data the solution will not work.

II. Ethical and Social Implications: There is an increasing number of ethical considerations coming out while deploying AI solutions. These range from biased dataset to unethical labor to copyright infringements and many other issues. From the very beginning it needs to be considered with proper policy and regulations.

III. Capacity Building and Skills Development: Building local capacity and skills in AI technologies is essential for sustainable development. Training stakeholders, including farmers, SMEs, and development practitioners, on AI tools and applications is crucial. Local ownership and adaptation of AI solutions ensure that they are contextually relevant and impactful.

IV: Sustainability: AI solutions are not cheap, there are concerns about uncomprehensible decision-making processes and the systems will require regular maintenance as well as upgrading to ensure security.

V: Less impact, more hype: As AI is the most used buzzword there are surely interests to use AI to make the project sound more interesting or tech savvy to the donors. During the blockchain hype, I have witnessed several blockchain projects that never needed blockchain in the first place. Or even earlier when farmers who used feature phones without internet were being added to whatsapp groups to disseminate agriculture information.

6. Final Worlds

I do not doubt that AI will become something as common as the internet. However, the expectation from AI needs to be rightsized. At present what we see as AI systems like ChatGPT, Copilot, Claude or Gemini all are weak AI models, and they are generating text, images, videos without understanding what it means truly (I am not going too technical as my intention is to keep the focus on MSD/VCD applications). As a result, what we can expect from AI depends on how we train the AI, what real world data we can provide and what precise instructions could be provided to the system to provide a result. International development is a very complex and complicated sector and there are many complicated human dynamics (history, emotions, culture, norm, politics, geo-politics etc.) that AI is unable to understand. ?Another very important factor is the relationship between data and AI; just as water is essential for fish, data is vital for the proper functioning of any AI system. And for many cases of MSD/VCD applications of AI the lack or quality of data would become an issue.

Finally, we are living in very interesting times and experiencing the slow but inevitable march of the fourth industrial revolution. Our approaches, thoughts and ideas need to incorporate the technological innovations that can boost the social innovations. MSD/VCD approach can take benefit of AI revolution or hype depending on how you see it, if we understand what we want, what resources are available and if we actually need AI for it. ??

#MSD #VCD #AI #artificialintelligence #marketsystems #systemicchange #digitaldevelopment #ICT4D #valuechaindevelopment #4IR