January 29, 2022
Kannan Subbiah
FCA | CISA | CGEIT | CCISO | GRC Consulting | Independent Director | Enterprise & Solution Architecture | Former Sr. VP & CTO of MF Utilities | BU Soft Tech | itTrident
Researchers also found additional hacking tools, from several sources, collected in the same repository. Alien Labs called the malware source code “simple yet efficient,” able to carry out malware attacks with a grand total of a mere 2,891 lines of code (including empty lines and comments). In its November writeup, Alien Labs noted that BotenaGo, written in Google’s open-source Golang programming language, could exploit 33 vulnerabilities for initial access. The malware is light, easy to use and powerful. BotenaGo’s 2,891 lines of code are all that’s needed for a malware attack, including, but not limited to, installing a reverse shell and a telnet loader used to create a backdoor to receive commands from its command-and-control (C2) operator. Caspi explained that BotenaGo has automatic setup of its 33 exploits, presenting an attacker a “ready state” to attack a vulnerable target and infect it with an appropriate payload based on target type or operating system. The source code leaked to GitHub and depicted below features a “supported” list of vendors and software used by BotenaGo to target its exploits at a slew of routers and IoT devices.
For starters, the "soft" skills will matter in the months and years ahead. These include professional skills such as communication, leadership, and teamwork, says Don Jones, vice president of developer skills at Pluralsight. Then there is a need for "tech-adjacent skills, like a familiarity with project management and business analysis." Jones urges an "evergreen" approach to skills mastery, as technology evolves too quickly to commit to a single platform or solution set. "The biggest-impact skill is the ability to learn," he says. "There's no single tech skill you can invest in that won't change or be outdated in a year; your single biggest skill needs to be the ability to update skills and learn new skills." This also means placing a greater emphasis on emotional intelligence, as many emerging systems will be built on artificial intelligence, analytics, or automation that mimic human processes, therefore augmenting human workers. "Anyone can be taught to swap out memory, but the skill of communication and responding to human emotion is not a skill so easily taught," says Chris Lepotakis
Web3 backers love to talk about how blockchain networks are computers that can be programmed to do anything you imagine, given superpowers by the fact that they are also decentralized. Ethereum was the first of these computers to get real traction, but it was quickly overwhelmed by traffic. Traffic is managed by charging fees to use the computer, and the fees to complete a single transaction on the Ethereum network can run over $100. Imagine spending $75 to create a “free” Facebook account and another $75 every time you wanted to post something, and you have a sense of what it would be like to participate in a social network on the blockchain today. Ethereum is in the midst of a transformation designed to make it more efficient — which is to say, faster, less expensive, and less wasteful of energy. In the meantime, technologists routinely appear announcing that they have built a more efficient blockchain. Solana, for example, is a company that raised $314 million last year to build what it calls “the fastest blockchain in the world.” With that in mind, let’s check in on how the fastest blockchain in the world was doing on Sunday, when the aforementioned crypto crash led many people to use it to buy and sell assets.
领英推荐
There is a growing challenge to better govern data as it increases in variety and volume, and there is an estimate that 7.5 septillion gigabytes of data is generated every single day. Moreover, in organizations, silos are getting created through multiple data lakes or data warehouses without the right guidelines, which will eventually be a challenge in managing this data growth. To achieve nimbleness, we can simplify the data landscape by using a semantic fabric, popularly called data fabric, based on a strong Metadata Management operating model. This can further make data interoperable between divisions and functions while working to a competitive advantage. Data fabric simplifies Data Management, across cloud and on-premise data sources, even though data is managed as domains. In addition, data democratization can be a strong enabler for managing data across domains with ease and making data available as well as interoperable. Allowing business users to source and consume relevant data for their instantaneous reporting or generation of insights can reduce significant turnaround time in acquiring or sourcing data traditionally.
The metaverse could potentially use virtual reality, or augmented reality as we know it now, to immerse users in an alternate world. The technology is still being developed, but companies like Meta say they are building and improving these devices. Meta's Oculus Quest, now in its second model, is one such device. "When you're in the metaverse, when you're in a virtual reality headset, you will feel like you're actually sitting in a room with someone else who can see you, who can see all of your nonverbal gestures, who you can respond to and mimic," Ratan said. Immersive worlds and creating online avatars is nothing new, as games like Grand Theft Auto Online, Minecraft and Roblox have already created virtual universes. Meta's announcement last October aims to go beyond entertainment, and create virtual workspaces, homes and experiences for all ages. "What's happening now is the metaverse for social media without gaming," Ratan said. "The new metaverse is designed to support any type of social interaction, whether that's hanging out with your friends or having a business meeting."
Data drift represents how a target data set is different from a source data set. For time-series data (the most common form of data powering ML models), drift is a measure of the “distance” of data at two different instances in time. The key takeaway is that drift is a singular, or point, measure of the distance between two different data distributions. While drift is a point measure, stability is a longitudinal metric. We believe resilient models should be powered by data attributes that exhibit low drift over time — such models, by definition, would exhibit less drift-induced misbehavior. In order to manifest this property, drift over time, we introduce the notion of data stability. Stable data attributes drift little over time, whereas unstable data is the opposite. We provide additional details below. Consider two different attributes: the daily temperature distribution in NYC in November (TEMPNovNYC) and the distribution of the tare weights of aircraft at public airports (AIRKG). It is easy to see that TEMPNovNYC has lower drift than AIRKG; one would expect lesser variation between November temperatures at NYC across various years, than between the weights of aircrafts at two airports.