Big Data, AI & Blockchain
In my opinion, there are three emerging technologies that will, if they haven’t already, transform the telecom sector. These technologies are big data, the most mature technology, artificial intelligence (AI) that feeds off of big data, and the blockchain.
Big data transformed AI by providing the immense volume of data needed for AI to machine learn. However one of the limiting factors for companies in investing in big data/AI initiatives has been the cost of centralized storage due to the limited budget. Blockchain-based technology will likely make an impact in the cost of storing data due to the disintermediation of centralized storage providers. For a company, it will also create downward pricing pressure on SaaS suppliers as they need to move to decentralized storage providers to deal with the volumes of data.
Blockchains could help DSP's verify, execute, and record digital transactions, and with AI they could monetize the data and improve the decision making. And with big data one can feed AI the volume of data it needs to learn and build models. Below are the characteristics of blockchain that can provide opportunities for AI :
- Decentralized and shared control: AI absolutely loves data, the more data it has to learn from, the better the which can help the DSPs achieve higher profits, lower costs or other relative business goals. This would mean mean an expansion of the concept of big data, moving from proprietary silos solutions to blockchain-enabled shared data layers. Data sharing might occur across an enterprise, consortium or a public blockchain. For example, an enterprise can build AI models that predict customer churn; a consortium of CSPs could make better payment models to prevent mobile money fraud.
- Immutability and Audit Trail: Blockchain can help solve the problem companies have with data management, a bad data that can come from an error , malicious act or due to a defective sensor. At each step of the process in building the AI model, the creator of the data can timestamps that model to the blockchain database, by including a digital signature saying it is valid. This would results in more trustworthy predictions.
- Security and Trust: Once you create data that can be used for model-building, you can restrict how others use them upstream as blockchain makes this easier for all use cases by treating permissions as assets. A user can grant read/write permissions to view a particular slice of data or model. For e.g. Google’s DeepMind’s healthcare AI which uses blockchain. A user simply owns their medical data and controls its upstream usage, therefore not putting the company at risk of regulation and antitrust issues.
Emerging technology like these definitely provides a tremendous opportunity for the DSPs to be able to provide data analytic insights with more trust than ever. Blockchain and AI will transform enterprise data strategies for a company. It is imperative for a DSP to not only understand each of the technologies bigdata/AI, including the role blockchain plays, but also to be a futurist thinker and not to be spending too much time competing but focusing on dominating the sector.
What's your views ?
