This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy

Optimizing high-speed networks for low power consumption is a challenge
07

Access, energy, literacy: overcoming the three stumbling blocks to connectivity

Telecoms entrepreneur and author Vijaya Kumar Ivaturi (IVK for short) is co-founder and CTO of Crayon Data, a big data and AI start-up based in Singapore.

What role do you think technology, and particularly ICT, plays in addressing global poverty?

Technology is often used to provide access, and to enable rapid scaling up. Most of the world is connected, and the rest - which is generally poor - also wants to be connected for services or benefits. Technology can bridge both the access and the distribution divides – and when we say tech we can assume it means digital.  


What are the obstacles to providing access and scaling up rapidly?

It's a good question. Access means a digital device – a mobile phone in 99% of cases - so one needs access to power for the phone charger etc. So, energy availability – as an enabler, a kind of table stakes for digital access - whether it is renewable or otherwise, is one of the major challenges. Network connectivity is the second issue. India is a good example in that there are many areas where you may have access to power, but the network access is spotty. So, for service delivery, fine tuning or customizing is required. The third issue is digital literacy. You can consume media and watch a video for sure. But if you want to access a service, to operate a bank account, or to pay somebody, it is not so trivial if you lack digital literacy. There are other economic reasons, such as affordability, income level etc., but these are the three from an enabling perspective that you typically encounter in the initial years. It has got a lot better, but it's not fully solved. 


On all three of those: energy access, network connectivity and literacy levels, do you think the rate of improvement is sufficient? 

We are a little better off on energy, because you can now be off-grid and still get power. It may not give you peak load performance, but you still have many options, using solar for instance, to get basic access. Wireless has solved some of the problems with network connectivity, but the difficulty is that a lot of the advances in wireless technology standards, whether 4G, 5G or 6G, essentially relate to data speeds and very high bandwidth. And that’s very power hungry. Optimizing such networks for very low power consumption is a challenge. That is something we need to keep in mind. Finally, the literacy issue is less technical and more social, cultural, and maybe even political. For countries like India, English is not the main language in rural areas, and you also have different scripts and dialects. Voice-enabled technologies and prompting solves many of these problems, but there’s still a lot of work to be done in changing the literacy level and making it more digital friendly.


You advise governments on these issues. Are they aware that addressing these factors is a key way to tackle poverty? 

The view is that to lift everybody out of poverty levels, you have to increase their income, which means you have to create more jobs and opportunities, and also give them access to digital services. You can’t solve the jobs creation problem through only public sector employment. The micro, small and medium enterprises (MSME) segment is where most of the entrepreneurship stays, and the thinking is that by encouraging them via funding and taxation policy it will automatically result in more jobs. 

That's true for some parts of the MSME market, but a lot of advanced technology doesn't necessarily mean more jobs. If you look at classical economics modeling, agri-jobs are typically the major percentage of the labor force, then you come to services and manufacturing. Services tend to be split into the high-level advanced technology jobs, which are normally garnered by well-educated, literate people, and then there are sub-optimal jobs. So, on paper you do have jobs, but they are not necessarily the same across the sectors. China and Japan are less focused on services and more on manufacturing. India, where services are the bigger segment, would like to improve manufacturing, because that has a better wealth distribution than purely high-tech services. If you have a social objective, increasing the manufacturing percentage of GDP will be better value. 

It’s a similar argument for financial inclusion in some African countries: I need you to have a phone, and once you have a phone, I can get you employment in a call center, or a support service. Because if you don't have a phone and a bike, you are pretty much unemployed in this world. Financial inclusion uses digital technology and that results in social inclusion. So, it’s not about the state taking everything and distributing it uniformly - those models don't really work. What works is the direct benefit transfer of money into people's hands, once they’re authenticated and identified. And technology enables that.


What role can global ICT providers such as Huawei play in combatting poverty, beyond funding skills training?

Global ICT corporations are generally platform players. They provide the digital highway and, beyond connectivity, maybe provide the service infrastructure for you. The assumption made is the infrastructure requirement is exactly the same across countries. Now, that is generally true from a technology angle. But when you go to the software cap-ex, I think the verticalization or regionalization of the solution is where they can play a role. If you look at Large Language Models, a lot of effort is being done in India today, for instance, in the local languages and local datasets, because that's what’s needed to interpret a diagnostic report or annotation for lung cancer. You cannot really understand what they mean if you don't have the support for local datasets. 

But local datasets are controlled by the nation, especially in healthcare and finance, because there are data sensitive requirements. And this is where a global corporation needs to make sure that they have a local relevance to many of the solutions. Generally, the position taken by them is to provide the basic structure and then invest in companies which build local solutions based on their platform. But if I have to do something else before I can use the app, that is beyond the realm of many companies. If you look at computing – away from quantum computing - there is not much difference between the methods used by most of these big companies. There is not much algorithmic variation. What varies is the dataset: the people who curate regional data sets, either by sector or by region, which have more depth and more fine-grain context. The game has moved to datasets rather than algorithms. I would say that is something which would be a very useful thing for global corporations to do more of, maybe in partnership. 


The first sustainable development goal SDG1 aims to eradicate extreme poverty worldwide altogether by 2030. Are you hopeful about that?

The meter will move but whether we completely eradicate it in six years, I don’t think so. It is a lofty goal and that’s good. But eliminate extreme poverty completely? I’m not so sure. I think we will create more jobs. Maybe from new types of remote healthcare which are employing many people – not necessarily doctors, but single-task experts. And the cleantech part will move far ahead and then you will find many new jobs, so it will have an impact. However, I don't think it'd be 100% across the states of India. Delays will not just be down to conviction and the cost of doing it, but also from geopolitical risks and the climate risk. So it’s a good way to move forward but I would say that it may be a stretch target. 

I would also see it from a slightly more social science or social psychology angle. A lot of these choices and changes have to come from individuals themselves, independent of the force or capital the government may provide. The aspiration of a certain lifestyle drives consumption patterns. Even if I train people, they may not want to shift to another place to get the job. You'd be surprised that they would rather be unemployed in their current location than move somewhere else. The local ecosystem has dependency both ways. Everyone seems to have an American dream. But expectation management is required to moderate what lifestyle you aspire to. It is not just about copying what somebody did ten years ago. Aspirations are normally crafted and imposed by the previous generation, but each generation needs a reset, a reframing. So you might see a different pattern going forward, but it takes some time. 


So, it's not just about the technology. You’re saying don't ignore the people factor within all this?

Yes, otherwise you won't solve it. You can have everything, all the technology, but nobody will actually go anywhere.


Contact us! transform@huawei.com