Yazar: mustafa efe

Kapil Thangavelu’s journey into cloud resource management began with an unexpected detour. Originally considering a career as a history professor, he pivoted to software development after discovering Python in college. His early exposure to Python-based tools like Zope and Plone set him on a path that led through diverse technical landscapes, from GIS work at DARPA to managing ephemeral AWS workloads at the Library of Congress. Eventually, he found himself at Canonical, working alongside Mark Shuttleworth on Ubuntu Linux and leading the development of Juju, an open-source project designed to simplify cloud application management. By the time Thangavelu joined Capital…

Java’s role in AI development is gaining momentum, challenging Python’s long-standing dominance in the field. While Python remains the go-to language for AI due to its extensive libraries and ease of use, Java advocates argue that the nearly three-decade-old language is well-positioned to make significant strides in AI. With its strong presence in enterprise environments and ongoing enhancements for AI capabilities, Java is emerging as a viable contender in the space. Oracle, the primary steward of Java, highlights a “triple play advantage” that positions Java favorably for AI. This strategy focuses on integrating AI with cloud infrastructure, improving AI-native capabilities…

Concerns over C++ memory safety may soon become a thing of the past. The C++ Alliance, a nonprofit organization dedicated to making C++ more accessible and secure, is spearheading efforts to introduce memory safety features through its Safe C++ Extensions proposal. This initiative aims to address longstanding issues with memory safety in C++ by providing developers with tools to write safer code without sacrificing the language’s flexibility and performance. If successful, this proposal could significantly enhance the security and reliability of software built with C++. The C++ Alliance plans to submit the Safe C++ Extensions proposal to the ISO for…

Java Development Kit (JDK) 23 has officially been released, bringing with it a host of exciting new features and improvements to the Java programming language. This update introduces a variety of changes, including previews of module import declarations, stream gatherers, structured concurrency, and scoped values, as well as a new class-file API. However, one significant change in JDK 23 is the deprecation of memory access methods in the sun.misc.Unsafe class, signaling a shift towards safer, more standardized alternatives. These updates aim to enhance the functionality and performance of Java, with thousands of smaller improvements and bug fixes included in the…

Generative AI heavily relies on data to generate responses to user queries. Large language models (LLMs), such as OpenAI’s GPT-3, are trained on enormous datasets to understand and produce natural language. For instance, GPT-3 was trained using the CommonCrawl dataset, which comprises 570 gigabytes of data and 400 billion tokens. These datasets, although vast, are essentially snapshots frozen in time, and they cannot incorporate real-time information about ongoing events. This limitation can lead to AI-generated responses that are outdated or, in some cases, even incorrect. Moreover, LLMs are susceptible to hallucinations—instances where the AI generates information that appears plausible but…

When it comes to AI, no executive is likely to give a free pass to experimentation without oversight. The reality for leaders is that AI efforts must be aligned with business goals, while carefully mitigating associated risks. Even the most cautious executives are hesitant to completely dismiss AI, knowing that failing to leverage AI-driven innovations could leave their businesses vulnerable to disruption and falling behind competitors. However, the balance between harnessing AI’s potential and managing its risks requires clear governance and accountability frameworks. AI governance encompasses the principles, practices, and regulations that guide organizations in utilizing AI effectively and responsibly.…

API development continues to experience tremendous growth year after year, and organizations are increasingly relying on APIs to connect services and facilitate data exchange. A report from 451 Research in 2022 found that the average organization is now managing over 15,500 APIs, with a staggering growth rate of 201% in just one year. Cloudflare’s data further highlights the prevalence of APIs, revealing that over half of the traffic it handles is API-based. However, as the number of APIs continues to rise, enterprises are facing significant challenges in managing and governing these various integration points, often leaving them fragmented and unregulated.…

We are still in the early stages of AI’s evolution, and one clear sign of this is how much work users still have to do to make AI tools function effectively. As Jono Bacon, founder of Community Leadership Core, points out, even something as simple as choosing the right large language model (LLM) to run a query can be confusing for most people. Once you’ve selected the model, there’s still plenty of work left to fine-tune the results, and consistency is far from guaranteed. Current AI models require a lot of manual intervention to ensure the output is relevant or…

A vector database may seem like just another type of database at first glance, but its functionality goes far beyond the traditional database model, particularly in the realm of artificial intelligence (AI). While conventional databases are optimized for handling structured, transactional data with relational queries, vector databases are designed to manage unstructured data, catering to modern AI-driven workloads such as machine learning inference, natural language processing, and recommendation systems. The key difference lies in how data is represented and retrieved. Traditional databases are used to store data in tables with predefined schemas and structured queries, whereas vector databases are tailored…

On September 16, a powerful open letter from key figures in the JavaScript ecosystem called out Oracle for what they consider the abandonment of the JavaScript trademark. The signatories of the letter included Ryan Dahl, creator of Node.js and Deno, as well as Brendan Eich, the creator of JavaScript. Together, they are urging Oracle to release the JavaScript trademark into the public domain. By September 20, the petition had gathered more than 10,000 signatures, signaling strong support from the broader tech community for their cause. The letter, titled “Oracle, it’s time to free JavaScript,” argues that Oracle’s ownership of the…