Mobile technology is growing and evolving at an overwhelming pace. The amount of computational power available in your pocket at this very moment is more than what the Apollo 11 spacecraft had onboard in the late 1960’s.
And it took a man to the moon.
So, with that in mind, don’t you think it’s useful to take a step back and realise exactly what your smartphone allows you to do at this present moment.
Maybe I’m showing my age, but it still feels like a scene from a science fiction movie when a voice from my phone starts guiding me to a specific location. And that’s only one aspect of mobile technology.
We need to start imagining what more we could do with it.
Last year Deloitte’s predicted, that in 2017, three hundred million smartphones (that’s a fifth of the predicted amount to be sold) would have some form of neural network machine learning capabilities onboard. That’s an enormous number of phones with remarkable capabilities.
Now, 2017 isn’t over yet, so there aren’t any figures that I could find to prove that is indeed happening. Nevertheless, it is at least a demonstration of the intent to get machine learning into our pockets.
And as always, Google, the great early-adopter, has already released Tensorflow Lite, an open source version of Tensorflow – their machine learning software – written specifically for Android devices. This enables Android developers to build and train neural networks on most mobile devices.
There are some big benefits to reap from all this activity. In the video below, Deloitte’s Arun Babu points out that the massive computing power in smartphones and their growing ubiquity means that people will be able to capitalise on machine learning to its fullest potential while not necessarily tethered to a desk.
This is great news for microlearning. This could lead to the production of more immersive applications all based on a mobile device. It could increase the reach of training initiatives throughout organisations to include employees who are often in the field, giving them the ability to tailor their training to suit their personal needs. This kind of training immersion could cascade through an organisation boosting staff retention and increasing communication within organisations.
Babu goes on to describe examples of possible uses for this machine learning and its usefulness to emergency response and disaster relief personnel on the ground. Whether controlling a drone flying over an area ravaged by fire or even predicting potential disasters, there are many practical settings that machine learning could be used in, for the betterment of all.
Access to consistent quality education has been a barrier to transforming the massive Sub-Saharan region of our continent into a major economic player – a multi-generational problem that has affected us for quite some time.
Mobile technology could be a solution to this problem. Schools with low infrastructure can look forward to radically increasing their access to information at a low cost. Many initiatives are already in position to assist them in maximising these great rewards.
Kenya, Côte d’Ivoire and various other countries, have all started digital schools to do exactly that. These schools equip learners with tablets and connectivity. This allows these low infrastructure schools to provide high-quality education to their learners.
Funded and supported by the Global e-Schools and Communities Initiative (GESCI) (that was founded by the United Nations Task Force on Information Communications Technology (ICT)), these schools utilise low-cost, yet robust information technology infrastructures and mobile devices, reducing the costs of implementation.
The bad news is that there are significant financial barriers to harnessing this potential – at least for now – and it is two-fold. The first is the price of a smartphone or similar device in South Africa and the rest of the continent, as it is still very high. However, there is an increasing number of affordable options on the market. Currently, STK and Hi-Sense both offer more affordable models.
The second financial barrier is the cost of data. And this is a sore point, . We have the second-highest data costs in the world – currently outranked only by Brazil. This is a very high hurdle that South Africa needs to jump, and the responsibility of solving this problem lies in a considered relationship between government and the private sector.
Mobile technology is becoming more affordable, more decentralised, more ubiquitous and more accessible; giving us better functionality and connectivity no matter how far away we may be from our desks. We should all take advantage of it as soon as we can.
References
Deloitte’s Arun Babu
Tensorflow Lite – deep machine learning on Android phones
South Africa’s Data Costs
South Africa’s Social Media Blackout
Gesci’s Digital Schools in Kenya and Côte d’Ivoire
Author: Simon Pienaar