Daily Tech Digest - January 16, 2024

Why Pre-Skilling, Not Reskilling, Is The Secret To Better Employment Pipelines

In a landscape where the relevance of skills evolves, Zaslavski says that organizations should focus on selecting and advancing individuals based on their potential for learning skills like critical thinking and resiliency, instead of focusing on hard skills like coding. ... “By concentrating on these fundamental elements, as opposed to current technical proficiency or past work history, organizations position themselves with an agile and future-ready workforce. In this light, pre-skilling should be an integral part of employers’ talent strategy pre and post-hiring, from sourcing and recruiting to career pathing and employee engagement.” ... She points to areas like understanding if a potential or existing employee has the EQ and social skills needed to perform as part of a group. Or whether they have the curiosity and analytical intelligence needed to learn new hard skills as well as the ambition and work ethic to achieve results. “When people have learning ability, drive, and people skills, they will probably develop new skills faster than others,” she says.


Agile is a concept we all continuously talk about, but what is it really?

Empiricism, teams, user stories, iterations; they are all examples of tools that we use in Agile, but they are not its purpose. Agile is about empowering people to take control of their environment and give them complete freedom to discover how to use available tools in the most effective way. And this applies to the why too. People adopt Agile to increase efficiency, transparency, velocity, predictability, quality. But again all these are a result of Agile, not its goal. It is the mindset that makes it all possible. That is why it is “People and interactions above processes and tools”. To illustrate this, think about empiricism itself. Try introducing empiricism into an organisation mired in a culture of fear and control, and it doesn’t work, no matter what you do. You can’t force empiricism. People are too busy evading blame and manipulating information. Think about it, how often do people complain that the retrospective doesn’t deliver anything? Retrospectives where people just complain and nothing changes? 


What Will It Take to Adopt Secure by Design Principles?

What does the future of secure by design adoption look like? CISA is continuing its work alongside industry partners. “Part of our strategy is to collect data on attacks and understand what that data is telling us about risk and impact and derive further best practices and work with companies, and really other nations, to adopt these principles,” Zabierek shares. International collaboration on secure by design is reflected not only in this CISA initiative but also the Guidelines for Secure AI System Development. CISA and the UK’s National Cyber Security Centre (NCSC) led the development of those guidelines, and 16 other countries have agreed to them. But like the Secure by Design initiative, this framework is also non-binding. A software manufacturer’s timeline for adopting secure by design principles will depend on its appetite, resources and the complexity of its products. But the more demand from government and consumers, the more likely adoption will happen. Right now, CISA has no plans to track adoption. “We're more focused on collaborating with industry so that we can understand best practices and recommend further better guidelines,” says Zabierek.


Mastering the art of motivation

Once you’ve helped employees connect their dots, the best way to further motivate them is also the cheapest, easiest, and has the fewest unintended consequences. Compliment them on a job well done, whenever they’ve done a job well enough to be worth noting. Sure, there are wrong ways to use compliments as motivators. First and foremost the employee you’re complimenting must value your opinion. If they don’t they’ll write off your compliment as just so much noise. Second, a compliment from you should not be an easy compliment to earn. “I really like your belt,” isn’t going to inspire someone to work inventively and late. Third, with few exceptions compliments should be public. There’s little reason for you to be embarrassed about being pleased with someone’s efforts. With one caveat: Usually you’ll have one or two in your organization who routinely perform exceptionally well, but also one or two who are plodders — good enough and steady enough to keep around; not good enough or steady enough to earn your praise. Find a way to compliment them in public anyway — perhaps because you prize their reliability and lack of temperament.


Do you need GPUs for generative AI systems?

GPUs greatly enhance performance, but they do so at a significant cost. Also, for those of you tracking carbon points, GPUs consume notable amounts of electricity and generate considerable heat. Do the performance gains justify the cost? CPUs are the most common type of processors in computers. They are everywhere, including in whatever you’re using to read this article. CPUs can perform a wide variety of tasks, and they have a smaller number of cores compared to GPUs. However, they have sophisticated control units and can execute a wide range of instructions. This versatility means they can handle AI workloads, such as use cases that need to leverage any kind of AI, including generative AI. CPUs can prototype new neural network architectures or test algorithms. They can be adequate for running smaller or less complex models. This is what many businesses are building right now (and will be for some time) and CPUs are sufficient for the use cases I’m currently hearing about. CPUs are more cost-effective in terms of initial investment and power consumption for smaller organizations or individuals who have limited resources. 


How to create an AI team and train your other workers

Building an genAI team requires a holistic approach, according to Jayaprakash Nair head of Machine Learning, AI and Visualization at Altimetrik, a digital engineering services provider. To reduce the risk of failure, organizations should begin by setting the foundation for quality data, establish “a single source of truth strategy,” and define business objectives. Building a team that includes diverse roles such as data scientists, machine learning engineers, data engineers, domain experts, project managers, and ethicists/legal advisors is also critical, he said. “Each role will contribute unique expertise and perspectives, which is essential for effective and responsible implementation,” Nair said. "Management must work to foster collaboration among these roles, help align each function with business goals, and also incorporate ethical and legal guidance to ensure that projects adhere to industry guidelines and regulations." ... It's also important to look for people who like learning new technology, have a good business sense, and understand how the technology can benefit the company.


Data is the missing piece of the AI puzzle. Here's how to fill the gap

Companies looking to make progress in AI, says Labovich, must "strike a balance and acknowledge the significant role of unstructured data in the advancement of gen AI." Sharma agrees with these sentiments: "It is not necessarily true that organizations must use gen AI on top of structured data to solve highly complex problems. Oftentimes the simplest applications can lead to the greatest savings in terms of efficiency." The wide variety of data that AI requires can be a vexing piece of the puzzle. For example, data at the edge is becoming a major source for large language models and repositories. "There will be significant growth of data at the edge as AI continues to evolve and organizations continue to innovate around their digital transformation to grow revenue and profits," says Bruce Kornfeld, chief marketing and product officer at StorMagic. Currently, he continues, "there is too much data in too many different formats, which is causing an influx of internal strife as companies struggle to determine what is business-critical versus what can be archived or removed from their data sets."


3 ways to combat rising OAuth SaaS attacks

At their core, OAuth integrations are cloud apps that can access data on behalf of a user, with a defined permission set. When a Microsoft 365 user installs a MailMerge app to their Word, for example, they have essentially created a service principal for the app and granted it an extensive permission set with read/write access, the ability to save and delete files, as well as the ability to access multiple documents to facilitate the mail merge. The organization needs to implement an application control process for OAuth apps and determine if the application, like in the example above, is approved or not. ... Security teams should view user security through two separate lenses. The first is the way they access the applications. Apps should be configured to require multi-factor authentication (MFA) and single sign-on (SSO). ... Automated tools should scan the logs and report whenever an OAuth-integrated application is acting suspiciously. For example, applications that display unusual access patterns or geographical abnormalities should be regarded as suspicious. 


Cloud cost optimisation: Strategies for managing cloud expenses and maximising ROI

Instead of employing manual resources, streamlining cloud optimisation through automation could bring enhanced resource savings to the table. The auto-scaling program offered by Amazon Web Services (AWS) is a shining example of how firms can effectively streamline their cloud optimisation in a short time. The program also enables swift optimisation in response to the changing resource requirements of systems and servers. ... At the planning stage, firms need to justify the cloud budget and ensure that unexpected spending is reduced to the minimum. The same approach has to be followed in the building, deployment, and control phases so that any unexpected rise in budgets can be adjusted promptly without throwing the entire financial control into a tizzy. All these steps will help organisations develop a culture of cost-conscious cloud adoption and help them perform optimally while keeping costs in check. ... Incorporating cloud cost optimisation tools is a strategic approach for organisations to streamline expenditures and enhance ROI. 


Pull Requests and Tech Debt

The biggest disadvantage of pull requests is understanding the context of the change, technical or business context: you see what has changed without necessarily explaining why the change occurred. Almost universally, engineers review pull requests in the browser and do their best to understand what’s happening, relying on their understanding of tech stack, architecture, business domains, etc. While some have the background necessary to mentally grasp the overall impact of the change, for others, it’s guesswork, assumptions, and leaps of faith….which only gets worse as the complexity and size of the pull request increases. [Recently a friend said he reviewed all pull requests in his IDE, greatly surprising me: first I’ve heard of such diligence. While noble, that thoroughness becomes a substantial time commitment unless that’s your primary responsibility. Only when absolutely necessary do I do this. Not sure how he pulls it off!] Other than those good samaritans, mostly what you’re doing is static code analysis: within the change in front of you, what has changed, and does it make sense? You can look for similar changes, emerging patterns that might drive refactoring, best practices, or others doing similar.



Quote for the day:

"All leadership takes place through the communication of ideas to the minds of others." -- Charles Cooley

No comments:

Post a Comment