There is no magic mixture for effecting digital transformation
When implementing digital transformation, expectations from digital technologies should be tempered. Terms like artificial intelligence (AI ), machine learning (ML), blockchain or the internet of things (IoT), after all, have to solve a real business or societal problem to make sense. For one, other than blockchain that was conceptualized a decade back, AI, ML, IoT, blockchain, 3D Printing, augmented reality, virtual reality, robotics and automation have been around for over three decades. So why this delayed euphoria?
AI’s growth is being driven by sophisticated algorithms that can be trained on mountains of data that requires a lot of computing power (bit.ly/2ztGuoD). ML, a subset of AI, does not require a computer to be explicitly programmed. Deep learning, an ML technique, is modelled on the human brain and uses artificial neural networks to feed inputs and get outputs. However, it still has to train on data and images to be effective—a good deep learning algorithm cannot be trained on bad data.
This brings us to the prowess of IoT. In this context, numbers like 50 billion devices by 2020 are thrown at us. That said, the fact is that data gleaned and analysed from cheaper sensors used in wearables, smartwatches and sectors like oil and gas, retail, healthcare and manufacturing are helping companies track damages, leaks and make better predictions—where to stock goods or monitor health remotely. IoT will get a fillip as sensors get smarter with AI and bandwidth speeds increase with 5G.
The younger blockchain, a technology that powers cryptocurrencies like bitcoin and ether, is witnessing private (permissioned) blockchain pilots and is being used by the financial and manufacturing sectors for smart contracts and improving supply chains among other uses. But it has to move beyond pilots to achieve success (bit.ly/2QBgoGV). 3D printing, as our cover story will also demonstrate, is being used to make everything from houses to aeroplane parts and even blueprints for 3D-printed guns (bit.ly/2P8idud).
Digital technologies, however, have challenges too— security, privacy, interoperability of standards, return on investment, and AI bias being some. Some of these issues are being addressed by the industry. For instance, “Explainable AI” is attempting to “enable human users to understand, appropriately trust, and effectively manage the emerging generation of AI partners (bit.ly/2x2sS3P)”.
Yet, there remains the looming fear of loss of jobs to AI, robotics and automation. Smart robots, for instance, are helping humans dispense with onerous and unsafe tasks. On the software side, robotic process automation also does away with repetitious work through automation. But this could result in loss of jobs.
In its the Future of Jobs Report 2018, the World Economic Forum estimates that by 2022, 75 million jobs may be displaced by a shift in the division of labour between humans and machines. On the positive side, 133 million new roles may emerge that are more adapted to the new division of labour between humans, machines and algorithms. However, the report adds that this “will require significant re- and upskilling”. Policymakers should take the cue and ensure that companies and governments do their part in making the transition smooth.
Source | Mint | 1st October 2018
Regards
Mr. Pralhad Jadhav
Master of Library & Information Science (NET Qualified)
Research Scholar (IGNOU)
Senior Manager @ Knowledge Repository
Khaitan & Co
Mobile @ 9665911593
No comments:
Post a Comment