Yazar: mustafa efe

JetBrains has announced its decision to discontinue several scripting-related technologies in the Kotlin language, including key components such as REPL functionality, Java scripting APIs, and certain Maven plugins. In a blog post dated November 19, JetBrains explained that the move comes after research showed that the user needs addressed by these technologies are already being fulfilled by other, more widely supported tools. The company’s focus is shifting to ensuring Kotlin remains aligned with evolving user requirements while streamlining its feature set. Among the changes, the default REPL (Read-Eval-Print Loop) implementations in both the Kotlin compiler and the IntelliJ IDEA plugin…

Anthropic has made a significant step toward advancing AI integration with data sources by introducing the Model Context Protocol (MCP), an open-source protocol designed to allow AI systems to seamlessly connect with various data sources. Unlike proprietary solutions that are limited to specific systems, MCP aims to create a universal interface that enables developers to build secure, two-way connections between AI tools and the data they need. This client-server architecture is intended to streamline the integration of AI with external information sources, making the process more standardized and accessible for a broad range of applications. The introduction of MCP addresses…

Building a simple prototype with ChatGPT might only take a weekend, but developing a fully operational generative AI system that can securely handle enterprise data is a whole different ball game. Enterprises often face significant engineering challenges when building production-ready AI systems. Development teams spend weeks or even months tackling issues like securing data pipelines, managing unstructured and structured data across siloed systems, configuring vector databases, choosing the right models, and implementing security measures—all while ensuring compliance with stringent industry standards. Traditional approaches to generative AI system development often require a tough decision: invest months of effort building custom infrastructure…

In the rapidly changing world of cloud computing, it’s clear that generic metrics are often inadequate for truly measuring success. While cloud units—metrics designed to standardize cloud costs and resources—can offer some insight, they’re not always a reliable gauge of cloud value. The problem is that many enterprises, even when they have cloud units in place, fail to create meaningful metrics that align directly with the business value they’re trying to achieve. In fact, many companies don’t use metrics at all, though they may not openly admit it. Cloud metrics need to be specifically aligned with the value they’re meant…

Deno Land, the organization behind the Deno runtime for JavaScript, has petitioned the United States Patent and Trademark Office (USPTO) to cancel Oracle’s trademark for JavaScript. The petition claims that Oracle has effectively abandoned the trademark, citing the fact that since acquiring it from Sun Microsystems in 2009, the company has not sold any JavaScript-related goods or provided services linked to the language. This move is part of a broader effort to challenge Oracle’s claim over the JavaScript name, which some argue has been left dormant for over a decade. The petition, submitted on November 22, follows a public plea…

It’s not surprising that many technology trends often mirror the patterns we see in the fashion world. While I’m not referring to our questionable dress sense as tech professionals, I’m talking about how we make decisions in the tech space. Right now, as you’re reading this, companies are throwing generative AI technologies like ChatGPT into their operations, hoping to achieve the same success that others have claimed. Commonwealth Bank of Australia, for instance, reports a significant reduction in scam losses and customer-reported fraud by using AI. While this is a victory for them, it’s important to realize that just because…

In the latest developments within the Python community and beyond, there’s plenty to explore. For those eager to dive into the newest version, Python 3.14 has just reached its alpha 2 release, and it’s already showing potential for exciting new features. Among the standout topics is the increasing popularity of standalone Python applications, with PyInstaller making it easier than ever to distribute Python apps as executables. Additionally, Python’s pattern matching feature continues to unlock new possibilities for developers, offering powerful tools for type-based decision-making. And for cloud enthusiasts, Microsoft Azure has launched a super-fast Python code sandboxing service, further boosting…

It may seem counterintuitive, but cloud computing projects today are facing more challenges and failures than they did a decade ago. Despite the rapid advancements in technology and infrastructure, current metrics reveal a concerning trend—cloud projects are not necessarily improving over time. A decade ago, most cloud initiatives were relatively straightforward, involving the migration of small test programs or systems. Today, however, the complexity of these projects has grown significantly, involving intricate systems that impact multiple facets of an enterprise. The growing integration of AI and data-intensive technologies has added another layer of complexity, making it harder for organizations to…

The announcement that a group of former Google executives is working on an operating system specifically designed for AI agents highlights a growing shift in how we conceptualize the role of AI in software platforms. Industry analyst Brian Jackson pointed out that the current operating systems we use are built around traditional file-based architectures and are designed for manual interaction via a computer and mouse. However, an AI-centric OS would redefine computing by focusing on data and continuous learning. This OS would harness intelligent models that adapt and evolve as they interact with new data, offering a dynamic, real-time response…

Uber is expanding its operations beyond rides and deliveries by launching a new division focused on AI data labeling. This initiative, known as Scaled Solutions, taps into the growing demand for data annotation, testing, and localization, offering gig workers a chance to support the development of AI models. Uber’s entry into this field is a strategic move to leverage its vast network of gig workers for more specialized tasks related to AI, catering to the needs of a diverse range of industries. Initially designed to meet Uber’s internal requirements for improving its AI systems, Scaled Solutions has evolved into a…