One of the greatest concerns in IoT is security, and how software engineers address it will play a deeper role. As devices interact with each other, businesses need to be able to securely handle the data deluge. There have already been many data breaches where smart devices have been the target, notably Osram, which was found to have vulnerabilities in its IoT lightbulbs, potentially gifting an attacker access to a user’s network and the devices connected to it. Security needs to be tackled at the start of the design phase, making requirement tradeoffs as needed, rather than adding as a mere ‘bolt on’. This is highly correlated to software robustness. It may take a little bit more time to design and build robust software upfront, but secure software is more reliable and easier to maintain in the long run. A study by CAST suggests that one-third of security problems are also robustness problems, a finding that is borne out in our field experience with customers. Despite software developers’ best intentions, management is always looking for shortcuts. In the IoT ecosystem, first to market is a huge competitive driver, so this could mean that security, quality and dependability are sacrificed for speed to release.
Often, journalists fixate on finding broken or abusive systems, but miss out on what happens next. Yet, in the majority of cases, little to no justice is found for the victims. At most, the faulty systems are unceremoniously taken out of circulation. So, why is it so hard to get justice and accountability when algorithms go wrong? The answer goes deep into the way society interacts with technology and exposes fundamental flaws in the way our entire legal system operates. “I suppose the preliminary question is: do you even know that you’ve been shafted?” says Karen Yeung, a professor and an expert in law and technology policy at the University of Birmingham. “There’s just a basic problem of total opacity that’s really difficult to contend with.” The ADCU, for example, had to take Uber and Ola to court in the Netherlands to try to gain access to more insight on how the company’s algorithms make automated decisions on everything from how much pay and deductions drivers receive, to whether or not they are fired. Even then, the court largely refused their request for information. Further, even if the details of systems are made public, that’s no guarantee people will be able to fully understand it either – and that includes those using the systems.
Data Mesh allows teams to curate/generate data and create usable data products for other teams. It also makes certain that platform teams can put their efforts into data engineering while data professionals can handle domain-specific data issues. While business data professionals are responsible for the quality and reliability of the data their teams produce, they can take assistance from platform teams in the face of technical glitches. Apart from that, Data Mesh design is inclined towards business users and requires relatively minor interference from platform teams. This is unlike centralized data teams that are responsible for everything, from data frameworks and access to dealing with data-related requests. To conclude, Data Mesh or the decentralized architecture encourages each party to excel in their area of expertise. The platform teams need to focus on technology, engineering, and data pipelines, while the data professionals are accountable for ensuring data quality. This holistic approach ensures end-users can perform their tasks by leveraging data insights without investing time in acquiring the results of a custom request.
Never stop learning. The skills you mastered a few years ago may be no longer relevant today, which is why it’s important to be open to constantly learning. Whether you are starting your career or have years of experience, take it upon yourself to learn new skills and technologies. ... The skills required to be a technologist have evolved, but also the ways with colleagues across lines of business. One change we’ve really embraced as an organization is embarking on an agile and product transformation. We’ve taken advantage of the opportunity that came with the changing behaviors of consumers over the past few years to really embrace agile at a different scale. This matters tremendously, because when we deploy code or build an entirely new product, it helps millions of consumers reach their financial goals. The pace of change has accelerated, but the focus on making it easier for our customers to bank with Chase has not. Today, we’ve reorganized ourselves away from project-based teams into product-based teams. Each product now has a dedicated tech, product, design, and data & analytics leader to help speed up decision making and improve connectivity and collaboration.
"It's a hugely important step Microsoft is taking to start blocking these macros by default, especially due to how invisible macros are to the majority of users," adds Nathan Wenzler, chief security strategist at Tenable, a vulnerability scanning company. "But that doesn't mean the threat is eradicated or we shouldn't continue to remind users to be vigilant about opening files from untrusted sources." Other companies are seeing threat actors switching tactics because of Microsoft's move, too. "The adversaries are aware of it," observes Tim Bandos, executive vice president of cybersecurity at Xcitium, a maker of an endpoint security suite. "They're testing out new ways of working around it because they're clearly not as successful now that Microsoft has made this change." Users of one notorious malicious program, known as Emotet, have already begun shifting tactics, he notes. "We've seen them shift recently from leveraging macros to using URLs to OneDrive or Google Drive," he says.
The consensus mechanism is a fundamental characteristic and differentiator among blockchains. Solana's consensus mechanism has several novel features, in particular the Proof of History algorithm, which enables faster processing time and lower transaction costs. How PoH works is not hard to grasp conceptually. It's a bit harder to understand how it improves processing time and transaction costs. The Solana whitepaper is a deep dive into the implementation details, but it can be easy to miss the forest for the trees. Conceptually, the Proof of History provides a way to cryptographically prove the passage of time and where events fall in that timeline. This consensus mechanism is used in tandem with another more conventional algorithm like the Proof of Work (PoW) or Proof of Stake (PoS). The Proof of History makes the Proof of Stake more efficient and resilient in Solana. You can think of PoH as a cryptographic clock. It timestamps transactions with a hash that guarantees where in time the transaction occurred as valid. This means the entire network can forget about verifying the temporal claims of nodes and defer reconciling the current state of the chain.
Introducing causality to machine learning can make the model outputs more robust, and prevent the types of errors described earlier. But what does this look like? How can we encode causality into a model? The exact approach depends on the question we are trying to answer and the type of data we have available. ... They trained the model to ask “if I treat this disease, which symptoms would go away?” and “if I don’t treat this disease, which symptoms would remain?”. They encoded these questions as two mathematical formulae. Using these questions brings in causality: if treating a disease causes symptoms to go away, then it’s a causal relationship. They compared their causal model with a model that only looked at correlations and found that it performed better — particularly for rarer diseases and more complex cases. Despite the great potential of machine learning, and the associated excitement, we must not forget our core statistical principles. We must go beyond correlation (association) to look at causation, and build this into our models.
To measure the success of an investment, you first need to quantify the cost of what you’re trying to protect. In a simplified model, the first step is to measure the given benefits of protection, this starts with an asset valuation. How valuable is this data to me? Those in charge of the budget need to execute the risk of that data not being protected. If I don’t take the necessary measures to mitigate the risk by investing in preventative cyber-security tools, how costly could this be when a breach occurs? It is more cost-effective to validate an organisation’s controls rather than spending money on more tools. By adopting specialised frameworks to counteract cyber threats, for instance, running a threat-informed defence, utilising automated platforms such as Breach-and-Attack Simulation (BAS), CISO’S can continuously test and validate their system. Similar to a fire drill, BAS can locate which controls are failing, allowing organisations to remediate the gaps in their defence, making them cyber ready before the attack occurs.
Beyond security issues, cloud outages can open the door to cascading disruptions affecting both routine business and mission-critical applications. “This can lead to [issues] ranging from revenue loss to more serious impacts -- such as putting lives at risk in the case of critical health care applications,” explains Ravikanth Ganta, a senior director at business consulting firm Capgemini Americas. A cloud outage’s seriousness hinges on several factors, including organization preparedness, the zone regions affected, and the services impacted. “In many cases, businesses that build and run their applications in the cloud can endure a cloud outage with little to no impact if they architect their applications to take advantage of the automated failover capabilities readily available in the cloud,” Potter notes. Modular applications designed to leverage loosely coupled services will typically experience only a minor drop in availability or performance during a vendor outage and, in many cases, may not be affected all. “Customers that ... haven’t architected their applications to gracefully failover or redirect traffic to unimpacted zones or regions, will face greater availability challenges when a cloud provider experiences an outage,” Potter says.
“A foundational aspect of DesignOps is the adoption of agile work breakdown structures (WBSs) to organize UX work from alignment with broad strategic objectives to screen-level details in a single EAP tool. While this feels foreign to most UX practitioners at first, agile WBS maps quite well to UX work. The business and operational benefits of this approach are profound, including more accurate plans, estimates, tracking and reporting.” With a single working environment for managers, designers, developers, and even stakeholders as part of the DesignOps strategy, everyone can easily align their work and tasks, test and comment on prototypes in real time, eliminate design handoffs, reduce costly iterations, keep track of progress and identify bottlenecks. ... There’s no such thing as a designer who can handle every process and task because in the end, they do everything but the actual design. Digital product design is a multi-layered job that requires different experienced units in particular fields. Just as there is a need for a separation between UX and UI design with two distinct experts handling each, there is a need for a dedicated DesignOps person.
Quote for the day:
"The task of the leader is to get his people from where they are to where they have not been." -- Henry A. Kissinger