Daily Tech Digest - March 22, 2022

When did Data Science Become Synonymous with Machine Learning?

Many folks just getting started with data science have an illusory idea of the field as a breeding ground where state-of-the-art machine learning algorithms are produced day after day, hour after hour, second after second. While it is true that getting to push out cool machine learning models is part of the work, it’s far from the only thing you’ll be doing as a data scientist.In reality, data science involves quite a bit of not-so-shiny grunt work to even make the available data corpus suitable for analysis. According to a Twitter poll conducted in 2019 by data scientist Vicki Boykis, fewer than 5% of respondents claimed to spend the majority of their time on ML models [1]. The largest percentage of data scientists said that most of their time was spent cleaning up the data to make it usable. ... Data science is a burgeoning field, and reducing it down to one concept is a misrepresentation which is at best false, and at worse dangerous. To excel in the field as a whole, it’s necessary to remove the pop-culture tunnel vision that seems to only notice machine learning. 


NaaS adoption will thrive despite migration challenges

The pandemic has also played a significant role in spurring NaaS adoption, Chambers says. "During the early days of COVID-19 there was a rapid push for users to be able to connect quickly, reliably, and securely from anywhere at any time," he says. "This required many companies to make hardware/software purchases and rapid implementations that accelerated an already noticeable increase in overall network complexity over the last several years." Unfortunately, many organizations faced serious challenges while trying to keep pace with suddenly essential changes. "Companies that need to quickly scale up or down their network infrastructure capabilities, or those that are on the cusp of major IT infrastructure lifecycle activity, have become prime NaaS-adoption candidates," Chambers says. It’s easiest for organizations to adopt small-scale NaaS offerings to gain an understanding of how to evaluate potential risk and rewards and determine overall alignment to their organization’s requirements.


Securing DevOps amid digital transformation

The process of requesting a certificate from a CA, receiving it, manually binding it to an endpoint, and self-managing it can be slow and lack visibility. Sometimes, DevOps teams avoid established quality practices by using less secure means of cryptography or issuing their own certificates from a self-created non-compliant PKI environment – putting their organizations at risk. However, PKI certificates from certified and accredited globally trusted CAs offer the best way for engineers to ensure security, identity and compliance of their containers and the code stored within them. A certificate management platform, which is built to scale and manages large volumes of PKI certificates, is perfect for the DevOps ethos and their environments. Organizations can now automate the request and installation of compliant certificates within continuous integration/continuous deployment (CI/CD) pipelines and applications to secure DevOps practices and support digital transformation. Outsourcing your PKI to a CA means developers have a single source to turn to for all certificate needs and are free to focus on core competencies. 


Reprogramming banking infrastructure to deliver innovation at speed

Fintech firms typically apply digital technology to processes those legacy institutions find difficult, time consuming, or costly to undertake, and they often focus on getting a single use case like payments, or alternative lending right. In contrast, neo banks, or challenger banks, deliver their services primarily through phone apps that often aim to do many things that a bank can do, including lending money and accepting deposits. A key advantage for both is that they don’t have to spend time, money, and organisational capital to transform into something new. They were born digital. Likewise, they both claim convenience as their prime value proposition. However, while customers want convenience, many still see banking as a high-touch service. If their bank has survived decades of consolidation and has served a family for generations, familiarity can be a bigger draw than convenience. That said, the COVID-19 pandemic has accelerated the online trend. More and more of us auto-pay our bills and buy our goods as well as our entertainment and services via e-commerce. 


No free lunch theorem in Quantum Computing

The no free lunch theorem entails that a machine learning algorithm’s average performance is dependent on the amount of data it has. “Industry-built quantum computers of modest size are now publicly accessible over the cloud. This raises the intriguing possibility of quantum-assisted machine learning, a paradigm that researchers suspect could be more powerful than traditional machine learning. Various architectures for quantum neural networks (QNNs) have been proposed and implemented. Some important results for quantum learning theory have already been obtained, particularly regarding the trainability and expressibility of QNNs for variational quantum algorithms. However, the scalability of QNNs (to scales that are classically inaccessible) remains an interesting open question,” the authors write. This also suggests a possibility that in order to model a quantum system, the amount of training data might also need to grow exponentially. This threatens to eliminate the edge quantum computing has over edge computing. The authors have discovered a method to eliminate the potential overhead via a newfound quantum version of the no free lunch theorem.


IT Talent Shortage: How to Put AI Scouting Systems to Work

The most likely people to leave a company are highly skilled employees who are in high demand (e.g., IT). Employees who feel they are underutilized and who want to advance their careers, and employees who are looking for work that can more easily balance with their personal lives, are also more likely to leave. It’s also common knowledge that IT employees change jobs often, and that IT departments don't do a great job retaining them for the long haul. HR AI can help prevent attrition if you provide it with internal employee and departmental data that it can assess your employees, their talents and their needs based upon the search criteria that you give it. For instance, you can build a corporate employee database that goes beyond IT, and that lists all the relevant skills and work experiences that employees across a broad spectrum of the company possess. Using this method, you might identify an employee who is working in accounting, but who has an IT background, enjoys data analytics, and wants to explore a career change. Or you could identify a junior member of IT who is a strong communicator and can connect with end users in the business. 


Automation and digital transformation: 3 ways they go together

All sorts of automation gets devised and implemented for specific purposes or tasks, sometimes for refreshingly simple reasons, like “Automating this makes our system more resilient, and automating this makes my job better.” This is the type of step-by-step automation long done by sysadmins and other operations-focused IT pros; it’s also common in DevOps and site reliability engineering (SRE) roles. IT automation happens for perfectly good reasons on its own, and it has now spread deep and wide in most, if not all, of the traditional branches of the IT family tree: development, operations, security, testing/QA, data management and analytics – you get the idea. None of this needs to be tethered to a digital transformation initiative; the benefits of a finely tuned CI/CD pipeline or security automation can be both the means and the end. There’s no such thing as digital transformation without automation, however. This claim may involve some slight exaggeration, and reasonable people can disagree. But digital transformation of the ambitious sort that most Fortune 500 boardrooms are now deeply invested in requires (among other things) a massive technology lever to accomplish, and that lever is automation.


The best way to lead in uncertain times may be to throw out the playbook

Organizations also used the sensing-responding-adapting model to combat misinformation and confusion about masks and vaccines. With conflicting guidance from the Centers for Disease Control and Prevention (CDC) in the US and the World Health Organization (WHO), one organization we studied opted for “full transparency” with a “fully digital” solution. The company built an app that included data from sources the company considered reliable, and it updated policies, outlined precautions, and offered ways to report vaccination status. The app turbocharged the company’s sense-respond-adapt capabilities by getting quality information in everyone’s hands and opening a new channel for regular two-way communication. There was no waiting for an “all hands” meeting to get meaningful questions and feedback. Reflecting on the results of the study, one takeaway became clear: it’s worthwhile for leaders of any team to absorb the lessons of sense-respond-adapt, even if there is no emergency at hand. Here are three ways to employ each step of the model.


The Three Building Blocks of Data Science

Data is worthless without the context for understanding it properly — context which can only be obtained by a domain expert: someone who understands the field where the data stems from and can thus provide the perspectives needed to interpret it correctly. Let’s consider a toy example to illustrate this. Imagine we collect data from a bunch of different golf games from recent years of the PGA Tour. We obtain all the data, we process and organize it, we analyze it, and we confidently publish our findings, having triple-checked all our formulas and computations. And then, we become laughingstocks of the media. Why? Well, since none of us has ever actually played golf, we didn’t realize that lower scores correspond to a better performance. As a result, all our analyses were based on the reverse, and therefore incorrect. This is obviously an exaggeration, but it gets the point across. Data only makes sense in context, and so it is essential to consult with a domain expert before attempting to draw any conclusions.


Surprise! The metaverse could be great news for the enterprise edge

Metaverse latency control is more than just edge computing, it’s also edge connectivity, meaning consumer broadband. Faster broadband offers lower latency, but there’s more to latency control than just speed. You need to minimize the handling, the number of hops or devices between the user who’s pushing an avatar around a metaverse, and the software that understands what that means to what the user “sees” and what others see as well. Think fiber and cable TV, and a fast path between the user and the nearest edge, which is likely to be in a nearby major metro area. And think “everywhere” because, while the metaverse may be nowhere in a strict reality sense, it’s everywhere that social-media humans are, which is everywhere. Low latency, high-speed, universal consumer broadband? All the potential ad revenue for the metaverse is suddenly targeting that goal. As it’s achieved, the average social-media junkie could well end up with 50 or 100 Mbps or even a gigabit of low-latency bandwidth. There are corporate headquarters who don’t have it that good. 



Quote for the day:

"The ability to summon positive emotions during periods of intense stress lies at the heart of effective leadership." -- Jim Loehr

No comments:

Post a Comment