Daily Tech Digest - December 20, 2020

What Is a Minimum Viable AI Product?

Most organizations don’t want to use a separate AI application, so a new solution should allow easy integration with existing systems of record, typically through an application programming interface. This allows AI solutions to plug into existing data records and combine with transactional systems, reducing the need for behavior change. Zylotech, another Glasswing company, applies this principle to its self-learning B2B customer data platform. The company integrates client data across existing platforms; enriches it with a proprietary data set about what clients have browsed and bought elsewhere; and provides intelligent insights and recommendations about next best actions for clients’ marketing, sales, data, and customer teams. It is designed specifically to directly complement clients’ existing software suites, minimizing adoption friction. Another integration example is Verusen, an inventory optimization platform also in the Glasswing portfolio. Given the existence of large, entrenched enterprise resource planning players in the market, it was essential for the platform to integrate with such systems. It gathers existing inventory data and provides its AI-generated recommendations on how to connect disparate data and forecast future inventory needs without requiring significant user behavior change.

Half of 4 Million Public Docker Hub Images Found to Have Critical Vulnerabilities

A recent analysis of around 4 million Docker Hub images by cyber security firm Prevasio found that 51% of the images had exploitable vulnerabilities. A large number of these were cryptocurrency miners, both open and hidden, and 6432 of the images had malware. Prevasio’s team performed both static and dynamic analysis of the images. Static scanning includes dependency analysis, which checks the dependency graph of the software present in the image for published vulnerabilities. In addition to this, Prevasio's team also performed dynamic scanning - i.e. running containers from the images and monitoring their runtime behaviour. The report groups images into vulnerable ones as well as malicious ones. Almost 51% of the images had critical vulnerabilities that could be exploited, and 68% of images were vulnerable in various degrees. 0.16%, or 6432 of the analyzed images had malicious software in them. Windows images, which accounted for 1% of the total, and images without tags, were excluded from the analysis. Earlier this year, Aqua Security’s cyber-security team uncovered a new technique where attackers were building malicious images directly on misconfigured hosts.

Getting Started— A Coder’s Guide to Neural Networks

The world is talking so much about machine learning and AI, but hardly anyone seems to know how it works, but then, on the flip side, everyone makes it seem like they’re experts on it. The unfortunate truth is that the knowledge and know-how seem to be stuck with the academic elites. For the most part, the material online for learning about machine learning and deep learning falls into 1 of 3 categories: shallow tutorials with barely any explanation on why certain patterns are followed; copy and paste material by those who want to pretend to have a self-made portfolio; or such intimidating math heavy lessons, that you get lost in all the Greek. This book was written to get away from all of that. It’s meant to be a very easy read which walks the reader through a journey on the fundamentals of neural networks. This books purpose is to get the knowledge out of the hands of the few and bring it into the hands of any coder. Before continuing, let’s clear something up. From an outsider’s perspective, the world of AI consists of so many terms which seem to mean the same thing. Machine learning, deep learning, artificial intelligence, neural networks. Why are there so many seemingly synonymous terms? Let’s take a look at the diagram below.

Here's how opinions on the impact of artificial intelligence differ around the world

Views of AI are generally positive among the Asian publics surveyed: About two-thirds or more in Singapore (72%), South Korea (69%), India (67%), Taiwan (66%) and Japan (65%) say AI has been a good thing for society. Many places in Asia have emerged as world leaders in AI. Most other places surveyed fall short of a majority saying AI has been good for society. In France, for example, views are particularly negative: Just 37% say AI has been good for society, compared with 47% who say it has been bad for society. In the U.S. and UK, about as many say it has been a good thing for society as a bad thing. By contrast, Sweden and Spain are among a handful of places outside of the Asia-Pacific region where a majority (60%) views AI in a positive light. As with AI, Asian publics surveyed stand out for their relatively positive views of the impact of job automation. Many Asian publics have made major strides in the development of robotics and AI. The South Korean and Singaporean manufacturing industries, for instance, have the highest and second highest robot density of anywhere in the world.

Artificial Intelligence In The New Normal – Attitude Of Public In India

There have been several discussions in the society about AI posing a threat to humanity and our way of living and working. In our study, 42% of the public believe that the impact of AI on net new jobs created will depend on the industry, and on balance feel that, overall, more new jobs will be created than lost (Net score 1%). 63% of the public feel that humans will always be more intelligent that AI systems. One puzzling trend that emerges in the study is about how the youngsters perceive AI. Those in age category less than 40 (Net score -8%) are relatively less optimistic that net new jobs will be created as compared to those aged greater than 40 (Net score 14%). Further, respondents aged less than 40 are 3 times less confident than those aged more than 40 that human intelligence will not be overtaken by AI. What explains this apparent diffidence among the youth? Or I wonder if they are being more prescient than the others about reaching singularity! I believe there is a need for appropriate education and communication strategies for the youth in India about AI and its positive potential. The public in India demonstrate a sense of optimism about the future in the new normal and believe in science and technology to make their lives better.

Data governance in the FinTech sector: A growing need

The neobanking model is another FinTech model that has seen significant traction globally. In India, neobanks primarily operate in partnership with one or multiple banking partner(s). This leads to sharing of data between the two entities for multiple banking services provided to consumers. To ensure regulated usage and security of customer data shared by banks with neobanks and vice versa, proper data security and access guidelines would need to be in place. Other FinTech segments, including payments and WealthTech, also require strong DG frameworks to ensure compliance both within the organisation and across its partners. In recent times, the industry has seen the introduction of several data-related laws and regulations aimed at ensuring the privacy and security of an individual’s PII and sensitive data. Some of the key focus areas include data sharing, data usage, consent and an individual’s data rights. Hence, there is increasing pressure on companies to remain compliant while adopting rapidly evolving FinTech models. Considering the changing regulatory landscape and requirements, some FinTech companies have already performed readiness assessments and have started to adopt an enterprise DG framework that would help them ensure effective data management ...

Ten Essential Building Blocks of a Successful Enterprise Architecture Program

The danger (or maybe, in some cases, the opportunities) for EAs is that they may be expected to be conversant in any type of architecture. In other words, some organizations may only hire one EA and expect her to be able to do any kind of architecture work except that of licensed architects. If one considers that EA work could be very different in, say, government organizations compared to for profit or non-profit ones, then one could imagine specialized EAs (e.g., Government Enterprise Architect, Non-Profit, Conglomerate Architect, etc.) that requires specialized training and experience. In fact, there has been general recognition that doing EA in government can be quite different from in profit-driven enterprises and therefore special frameworks training for government-centric EA may be appropriate. Nonetheless, the leading generic, openly available EA framework for professional certification is The Open Group Architecture Framework (TOGAF), which, with expert assistance, can be adapted to incorporate elements of both DODAF and the FEA Framework (FEAF). With so many frameworks, methods, and standards to choose from, why is customization always required?

How Data Governance Can Improve Corporate Performance

While data governance is a systematic methodology for businesses to comply with external regulations such as GDPR, HIPAA, Sarbanes-Oxley, and future regulations, it can also establish a foundation and controls to strengthen internal decision-making for determining product costs, inventory, consumer demand, and more. While there are many factors to consider for building a data governance program, two of the most pressing items that should be top of mind are data quality and self-service analytics. It’s advantageous to include efforts to ensure data quality is part of your data governance program. Trying to govern data that is old, corrupted or duplicated can become quite messy. Although the tools for managing quality and governance are generally different, data governance provides a framework for data quality. Poor data quality exists for many reasons, such as having data spread out in department silos, different versions of the “same” data or information lacking in common name identifiers. Without data quality, organizations also face a real possibility of making faulty business decisions and having a sub-standard governance program. Generally, the more data governance a company has, the stronger its data quality will be.

5 Steps to Success in Data Governance Programs

What exactly does a successful data governance program look like? Author Bhansali (2014) defines data governance as “the systematic management of information to achieve objectives that are clearly aligned with and contribute to the organization’s objectives” (p.9). So, a successful data governance program is one that achieves these aligned objectives and furthers the interests of the organization to which it is applied. In our reading for this week (Bhansali, 2014) outlined several key steps in the creation of data governance platforms. These steps are by no means an exhaustive road-map for a perfect data governance platform, nor are they necessarily chronological. Still, they do provide a launching point for useful discussion. A data governance program must be aligned with any existing business strategies. This also involves being aware of the vision of the future that guides and defines the business. If Apple were the company under consideration, you might think of their vision being an iPhone in the pocket of every person on earth. Create a clear and logical model of the data governance process that is specific to your organization. This model should stand apart from any products or technologies created by the company and must be based on any key processes or standards

Application Level Encryption for Software Architects

Unless well-defined, the task for application-level encryption is frequently underestimated, poorly implemented, and results in haphazard architectural compromises when developers find out that integrating a cryptographic library or service is just the tip of the iceberg. Whoever is formally assigned with the job of implementing encryption-based data protection, faces thousands of pages of documentation on how to implement things better, but very little on how to design things correctly. Design exercises turn out to be a bumpy ride every time you don’t expect the need for design and have a sequence of ad-hoc decisions because you anticipated getting things done quickly: First, you face key model and cryptosystem choice challenges, which hide under “which library/tool should I use for this?” Hopefully, you chose a tool that fits your use-case security-wise, not the one with the most stars on GitHub. Hopefully, it contains only secure and modern cryptographic decisions. Hopefully, it will be compatible with other team’s choices when the encryption has to span several applications/platforms; Then you face key storage and access challenges: where to store the encryption keys, how to separate them from data, what are integration points where the components and data meet for encryption/decryption, what is the trust/risk level toward these components?; 

Quote for the day:

"No one reaches a high position without daring." -- Publilius Syrus

No comments:

Post a Comment