Latest news with #C++
Time of India
6 days ago
- Business
- Time of India
Job losses: How AI has painfully disrupted dreams of young software engineering graduates
IT CEOs have indicated that AI-led productivity is changing the business model, with revenue growth and headcount growth being de-linked. 'The last couple of years, we have been challenging our teams on how you can deliver twice the revenue and half of the people,' said HCLTech CEO C Vijayakumar in February. Tired of too many ads? Remove Ads Tired of too many ads? Remove Ads Tired of too many ads? Remove Ads ( Originally published on May 31, 2025 ) As a story going viral recently recounts, back in the early 1990s, Infosys cofounder Nandan Nilekani had prodded and pestered actor and playwright Girish Karnad, a distant relative, to buy into the then-obscure software firm's told to journalist Rollo Romig, author of I Am on the Hit List: Murder and Myth-making in South India, Karnad eventually gave in and bought some shares of Infosys. Within 10 years, as Infosys—and India's burgeoning IT sector—g rew, the share prices skyrocketed and helped Karnad out of a lower-middle class living to greater comforts, like a house of his There is no other word that encapsulates what C++, Java and Python did for India and millions of folks like Karnad. Beyond shareholders, zeros and ones carried with them the aspirations of millions of youth who gained not just employment, but a living that lifted their families out of the lower middle-class trap, powered by fancy salaries, lucrative stock options and promise of foreign far so good. Then, out of nowhere, came the threat from artificial intelligence (AI). India's middle-class dreams, written in the promise of software, is now under threat from advancements of that very jobs that millions of students had taken for granted as an entry to a long and successful career aren't quite there anymore, and a thirty-year dream is starting to lose drastic shift is leaving a bloody trail of laid-off employees, changing job descriptions and under-skilled young (name changed), a techie in Bengaluru, is job hunting. This isn't the best time to be looking for one. But he does not have a choice as his company, a unicorn, fired him five months ago, along with close to a dozen last time he was looking for a job was in 2018 when he was a final-year engineering student. Back then, all that the unicorns he was interviewing for wanted was solid programming six years, he was laid off and there are more things he is worried about than getting the basics right. 'Even if I get a job, how long can I hold on before the company decides otherwise? Is this going to be the end of my career?'Things are worse for junior developers just entering the workforce where AI tools can do a much better job. Their days are now marked by anxiety, fear and insecurity that threatens careers, lest they don't keep up with the change, and at times even when they (name changed), a manual tester in a Bengaluru-based IT services firm a decade ago, remembers how worried they were when automation was introduced. 'We were worried that our jobs would be lost,' she never came to pass in the five years she spent in the firm before moving to consulting. But today, testing is one of the areas seeing the most automation, and others such as frontand back-end development are soon likely to of this, naturally, is leading to mental health Singh Saluja, president of IT professionals' welfare association Nascent IT Employees Senate (NITES), has been seeing increasing anxiety in young professionals with up to five years of experience, who were beginning to feel they were being gradually sidelined or replaced.'Many are unsure whether their job will still exist in the next six months or a year,' Saluja says.'Every single project that you do, they track how many AI tools or AI integrations you are using. They don't always say it, but the bottom line is that if you don't, your job is at risk,' says long Reddit threads, software developers have been sharing how their department heads emphasise using AI tools and are removing teams that were doing documentation, something that has since been easily the moment it is each to their stay relevant, many are upskilling and learning AI-first thinking and how to create workflows using AI. Platforms such as Scaler Academy, Newton School and 100xEngineers are seeing huge demand for their online courses on AI and ML.'It is a six-month weekend course, which is a mix of lectures and hands-on exercises,' says Sridev Ramesh, cofounder, new-a g e schools profit, engineering colleges that mushroomed across the country over the last few decades are just not equipped for this transition, and that is resulting in students charting their own Rachit (last name withheld to protect identity), a second-year computer science student. He was clear that regular engineering colleges might not help. After preparing for IIT-JEE, he decided to pursue a four-year undergraduate degree with Newton School of Technology, which focuses on avid programmer from Class 8, he taught himself Java and then Python, and is currently interning at one of the top AI startups in India and in his words 'is loving it'.As Nishant Chandra, CEO, Newton School, points out, the ecosystem is changing fast and students need to change with it. Chandra reckons that unfortunately about 90% of the colleges are not forward-looking, and that will impact the (name changed), a third-year engineering student, and his batchmates often discuss what AI would do to their prospects. 'We are still a year from when we have to face it, but at present, we are unsure what we can do,' he who hails from a tier-3 town in Kerala, is doing computer science in Coimbatore. Ask him if the college is taking additional initiatives to equip them, and he is confused. 'We have not heard anything from the college. Maybe we will see something before we start placements next year,' he by the time reality hits, it might be too late for students like Sharma, CEO, TeamLease Digital, says most engineering graduates are not completely ready for AI jobs.'More than 60% of these students don't have enough hands-on knowledge and experience,' says Sharma, adding that beyond college degrees, what's needed is certifications in AI, cloud, security, or data science, working on real projects (like sharing code on GitHub), and joining hackathons or inter nships. 'Students who keep learning and can show real projects or skills will have the best chance of getting hired in today's job market.'That's not going to be easy.'We can't just learn one or two skills and assume that it will take us through the next five years,' says Savita Hor tikar, global head of talent acquisition at the AI company Fractal, adding that adapting to the new reality of 'continuous learning' is often harder for experienced professionals than CEOs have indicated that AI-led productivity is changing the business model, with revenue growth and headcount growth being de-linked. 'The last couple of years, we have been challenging our teams on how you can deliver twice the revenue and half of the people,' said HCLTech CEO C Vijayakumar in means AI taking over grunt work and humans focusing on strategy, ethics and innovation, says Roop Kaistha, regional managing director-APAC at recruitment firm is going to complicate is home to the second largest pool of software developers in the world, with 5.8 million professionals. It also produces around 1.5 million fresh engineers every year. However, just 10% of them have the ability to secure jobs, according to a TeamLease the question of higher-level output is going to be a pipe dream unless both the individual and the system change their companies and institutions need to work together and create courses that match what businesses actually use now, says TeamLease's Pai, director of Takshashila Institution, a centre for research and education on public policy, says that as long as companies and workers are prepared to learn, adapt and adjust, India will benefit from the AI revolution just as it has benefited from previous turns of the tech Sherikar, head of corporate development, Sonata Software, says self-skilling has also become nonnegotiable with AI-readiness now being a baseline expectation.'Skilling programmes must evolve from being theoretical to being outcome-focused, anchored in the realities of a tech-driven, rapidly changing business landscape,' she says.'While we have a number of skilling programmes underway, I think they are too disaggregated,' Sangeeta Gupta, SVP & chief strategy officer of Nasscom, recently told ET. 'You need a much more top-down thinking on skilling, not just for the top-end of AI, which is all the data scientists and that kind of work, but how will the workingage population be using AI more effectively in the day-to-day operations?'Until that happens, the pain will A Damodaran, professor, economics, IIM-Bangalore, says, 'We have seen automation disrupting businesses historically. The biggest was textiles and then factories, where automation led to job losses. In factories, before automation, many of the workers were handling hazardous materials, and nobody voluntarily did that. And as history would show, people found other jobs. That will happen again.'Unfortunately for the next generation of millions who were betting on software jobs, history will unfold far too slowly.
Business Insider
15-05-2025
- Business
- Business Insider
Why Citadel Securities is training its developers on a coding language update so complex it hasn't even been released yet
You may think that coding languages are static, just a string of letters and numbers for humans to communicate with machines and software. But Herb Sutter, a tech leader at Citadel Securities, says otherwise. "All the major languages that are in heavy use are living languages," Sutter told Business Insider. "That's why we see C++, Rust, C, and Python continuing to evolve. Our landscape is always changing, and it's important to stay abreast of those developments." As a market maker, Citadel Securities needs to be ready to match buyers and sellers and provide liquidity to institutional and retail investors worldwide. The company is focused on mastering C++, because speed and execution are everything. It's considered a more specialized coding language that is often used at high-frequency trading firms and exchanges. Better use and understanding of C++ can translate to faster systems and fewer coding mistakes. Sutter joined in 2024 from Microsoft to spearhead its training initiatives on C++, which is used extensively throughout Citadel Securities' technology. As one of the more senior technologists at the firm, it's Sutter's day job to keep up with the evolution of coding languages to make sure Ken Griffin's market maker is reaping the benefits of the latest and greatest. But even less experienced coders have something to gain by familiarizing themselves with the fresh features that come with new versions of C++; one edge is standing out in the interview process to nab a job at Citadel Securities, which can fetch up to $350,000 for jobs requiring C++ experience. In this Q&A, Sutter discusses how the firm is embracing a new version of C++ that isn't even set to be fully released until next year, and two pieces of advice that can help engineers stand out from the crowd. It has been edited for length and clarity. I've been at Citadel Securities for about six months. Can you believe it? And it has been great. I've been drinking from a firehose because there's lots of exciting work to do and new things I'm being exposed to. I've particularly enjoyed seeing how the firm is adopting the important and immediately useful elements of the new standards, even without waiting for the ink to officially dry. One of the things I'm especially excited about is C++'s async framework that's coming in the next standard that will ship about a year from now. Async use of C++ is a big deal because we are all increasingly needing to do things concurrently and in parallel. [Editor's note: "async" is shorthand for asynchronous — code that can run in the background without freezing your app. It's a new framework that helps developers write faster, smoother programs by handling tasks like downloading files or crunching numbers without making users wait.] I didn't realize until I joined Citadel Securities just how much that framework is already used at the firm, including for our US equities trading. Working at Citadel Securities is almost like living in the future in that we're already diving deep into technologies that will eventually be widely used. That's been a lot of fun. Concurrency is the idea of doing more than one thing at a time, which we're always trying to do in a network-cloud world, whether that's waiting for cloud capacity or AI tokens. Parallelism is when you have one huge computation to do, but would like to spread elements out over multiple machines to get the answer more quickly. What impresses me most about C++26's async framework is that it can handle both of those elements — the one that involves hiding and waiting, and the other that's doing many different things. Those are such different things. Doing them both well in one framework is pretty amazing. Just think of what a trading system has to do. Requests for trades are flying across the wire all the time. You never know when the client is going to want to buy or sell, so as a market maker, you have to be ready at all times. And that means being very responsive, very efficient, and super fast. Execution is extremely important, and that's why we're investing in the async framework. What are some of the advantages that you're seeing being an early adopter or a first mover in this new C++ standard? At Citadel Securities, using the things today that everyone else is going to be using months or years from now builds muscle and familiarity — especially for something as core as an async framework. Beyond that, we have been providing feedback and suggesting tweaks to the standard that are being adopted. Because we're using the new standard in production and at scale, we're able to play a role in evolving the language, which has been great. One way you can show off your C++ skills is simply by talking about what you're looking forward to most in C++26, describing the features that have helped you, or sharing something you've learned recently. I want to know that you're that curious and that you're focused on continuous learning, and that's true more generally, regardless of language. It's important to be able to demonstrate curiosity and knowledge about software advancements — and to show that you understand that there's more than one tool out there and that you know how to use them together. What's your advice to young engineers interested in joining Citadel Securities? I would encourage young engineers to get as much work experience as possible as early as possible. I went to the University of Waterloo in Canada, which has a well-known co-op program that served me incredibly well, but there are many others out there. The key is to get some work experience so that by the time you graduate, you have spent a significant amount of time using your skills in a real-world environment. The biggest differentiator we see among candidates is their ability to use technology to solve commercial problems. Ultimately, knowing data structures, languages, and the like are important tools in the toolkit, but what will really set you apart is your ability to solve business problems.
Time Business News
04-05-2025
- Time Business News
Mastering the Python Compiler: A Complete Guide for Beginners and Developers
Python has become one of the most popular programming languages due to its simplicity, versatility, and wide-ranging applications—from web development to artificial intelligence. However, while writing Python code is relatively straightforward, many beginners often overlook an important part of the programming process: how the Python compiler works. Understanding the role of the compiler in Python can significantly enhance your programming efficiency, help you debug issues, and optimize your code for better performance. This detailed guest post will guide you through everything you need to know about the Python compiler, its functions, types, and how to choose the right approach for your coding journey. Before diving into Python-specific details, it is essential to understand the general concept of a compiler. A compiler is a program that translates code written in a high-level programming language (such as Python, C++, or Java) into machine code that a computer's processor can execute. This process usually involves several stages: Lexical analysis : Breaking the source code into tokens. : Breaking the source code into tokens. Syntax analysis : Checking for syntax errors. : Checking for syntax errors. Semantic analysis : Ensuring logic is correct. : Ensuring logic is correct. Code generation : Producing machine-level code. : Producing machine-level code. Optimization: Improving performance and reducing resource usage. However, Python's approach to compilation is slightly different from traditional compiled languages. Unlike C or C++, Python does not compile directly to machine code. Instead, Python uses an interpreter, but it also compiles source code into bytecode before interpreting it. Step Description Source Code (.py) The human-readable code written by the developer. Bytecode (.pyc) Intermediate, platform-independent representation compiled by the Python compiler. Python Virtual Machine (PVM) Executes bytecode and returns results. So, Python is both a compiled and interpreted language. The compiler compiles the source into bytecode, and the interpreter executes it. Using bytecode offers multiple advantages: Portability : Bytecode is platform-independent, meaning the same code runs on Windows, Linux, or macOS without modification. : Bytecode is platform-independent, meaning the same code runs on Windows, Linux, or macOS without modification. Performance : Bytecode runs faster than raw source code. : Bytecode runs faster than raw source code. Security: Bytecode can be obfuscated to prevent unauthorized code reading. It is essential to distinguish between Python's compiler and interpreter functions. Aspect Compiler (Python Compiler) Interpreter (Python Virtual Machine) Role Translates source code to bytecode Executes bytecode Speed Happens once during compilation Runs every time during execution Output Bytecode (.pyc files) Program output or runtime results Use case Improving performance, caching Running the application Both work together seamlessly to make Python efficient and easy to use. Several Python compilers and implementations exist, each serving different needs: Python Compiler Description CPython The default and most widely used compiler. Compiles Python code into bytecode and interprets it. PyPy An alternative with Just-In-Time (JIT) compilation, significantly improving execution speed. Cython Translates Python code to C for faster execution and allows static typing. Jython Compiles Python code into Java bytecode for JVM execution. IronPython Compiles Python code for .NET framework applications. For most developers, CPython is sufficient. However, PyPy and Cython are excellent choices for performance-critical applications. Python compilers are crucial in various scenarios: Performance Optimization : PyPy or Cython can speed up CPU-heavy Python applications. : PyPy or Cython can speed up CPU-heavy Python applications. Cross-Platform Compatibility : CPython ensures code runs across multiple operating systems. : CPython ensures code runs across multiple operating systems. Integration with Other Platforms : Jython and IronPython allow Python to integrate with Java and .NET ecosystems. : Jython and IronPython allow Python to integrate with Java and .NET ecosystems. Code Distribution: Bytecode makes distributing Python applications easier and protects source code. Though Python automatically compiles scripts to bytecode upon execution, you can manually compile Python files using built-in modules: python -m compileall This command generates .pyc files, which are the compiled bytecode versions of the source files. Faster Startup : Precompiled bytecode can speed up application startup time. : Precompiled bytecode can speed up application startup time. Code Protection : Distributing .pyc files instead of .py files helps obscure source code. : Distributing .pyc files instead of .py files helps obscure source code. Reduced Runtime Errors: Compilation can catch syntax errors before runtime. Though Python handles compilation automatically, developers may face some issues: Issue Possible Reason Solution SyntaxError during compilation Invalid Python syntax Check for typos, indentation errors Incompatible bytecode (.pyc) Running on different Python versions Recompile on the correct version ImportError due to missing compiled files Missing .pyc files Manually compile using compileall Understanding how Python compiles and executes code makes troubleshooting significantly easier. To maximize the efficiency of Python compilation: Keep Python Up-to-Date : Use the latest version for optimizations and security updates. : Use the latest version for optimizations and security updates. Leverage PyPy or Cython for Speed : Consider alternatives when execution speed is crucial. : Consider alternatives when execution speed is crucial. Use Virtual Environments : Prevent conflicts between Python versions and dependencies. : Prevent conflicts between Python versions and dependencies. Precompile for Deployment: Reduce load time and protect source code by distributing .pyc files. While Python's compilation model has stayed consistent, improvements are continuously being made: Static Typing with Type Hinting : Improves the possibility of further optimizations during compilation. : Improves the possibility of further optimizations during compilation. JIT Compilation in CPython (PEP 659) : Recent Python versions are exploring JIT to improve performance. : Recent Python versions are exploring JIT to improve performance. Better Bytecode Optimization: Future compilers may generate more efficient bytecode, reducing execution time. As Python continues to grow, its compiler and compilation processes will likely become even more sophisticated and powerful. Understanding the Python compiler is vital for both beginners and advanced Python developers. Whether you are writing simple scripts or developing complex data-driven applications, knowing how your code is compiled and executed will help you: Improve your program's performance. Reduce bugs and errors. Optimize for deployment and scalability. From CPython's standard compilation to advanced compilers like PyPy and Cython, Python offers flexibility to suit various needs. As Python evolves with features like JIT compilation and enhanced type hinting, mastering the Python compiler will remain a valuable skill for anyone serious about programming with Python. By gaining a deeper understanding of Python's compilation process, developers can bridge the gap between writing code and delivering efficient, reliable software solutions. TIME BUSINESS NEWS
Arabian Post
01-05-2025
- Business
- Arabian Post
Microsoft Embraces AI for Code Development, CEO Reveals
Microsoft CEO Satya Nadella has confirmed that between 20% and 30% of the company's code is now written by artificial intelligence , underlining a significant shift in how software is being developed. Speaking at Meta's LlamaCon conference, Nadella explained that AI's growing role in coding reflects broader trends in the tech industry, where machine learning models and advanced algorithms are streamlining development processes and augmenting human programmers' capabilities. The discussion at the conference also delved into comparisons between AI's performance in different programming languages. When asked whether AI is more adept at writing Python or C++, Nadella pointed to the unique characteristics of each language and how AI tools are evolving to handle both with increasing proficiency. Python, due to its simplicity and wide usage in AI and machine learning applications, has seen more extensive adoption by AI-driven coding assistants. However, C++, with its complexity and performance demands, presents a more challenging landscape for AI code generators, yet recent advancements show notable improvements in this area as well. Nadella's comments came during an interactive segment where he turned the spotlight on Mark Zuckerberg, CEO of Meta, asking him how much of Meta's code is being written by AI. The exchange highlighted the industry-wide interest in leveraging AI to augment and speed up development, prompting questions about the balance between human expertise and machine-generated output in software engineering. The conversation also sparked broader reflections on the future of AI in coding and software development. Nadella emphasised that while AI tools are making great strides in automating code generation, they are not intended to replace human developers but to assist them in increasing efficiency and tackling more complex problems. AI's role is particularly potent in repetitive tasks such as bug fixing, testing, and code optimisation, allowing developers to focus on higher-level, more creative challenges. This evolving trend of AI involvement in software development is not limited to Microsoft. Other major players in the tech sector are also exploring how AI can reduce the time and cost associated with writing code. With tools like GitHub Copilot, powered by OpenAI's models, developers are already leveraging AI to draft code snippets, suggest improvements, and enhance productivity. As these tools become more sophisticated, there is a growing push towards AI-driven development environments that assist in all stages of the software creation process. Meta, under Zuckerberg's leadership, has been at the forefront of AI innovation as well. The company's own research in AI and machine learning is contributing to the creation of LLaMA , an open-source model designed to push the boundaries of natural language processing. The focus on open-source collaboration is seen as a key strategy to accelerate the deployment of advanced AI tools in various industries, including software development. The role of AI in code generation is also shifting the conversation around job roles in the tech industry. While some have expressed concern that AI might replace jobs, others, including Nadella, believe the opposite is true: AI will empower developers to be more productive and creative. The ultimate goal is not to eliminate human involvement but to transform the way software is developed by automating mundane tasks and providing developers with powerful tools to enhance their capabilities. Despite the enthusiasm surrounding AI in coding, challenges remain. One of the primary hurdles is ensuring that the code generated by AI is accurate, secure, and free of bugs. As AI continues to evolve, developers must still closely monitor the output, refining and testing it before deployment. This creates an ongoing need for human oversight and intervention, underscoring the symbiotic relationship between AI and human developers. AI-driven coding tools have already had a profound impact on productivity. For example, automated refactoring tools powered by AI can restructure code for better efficiency or readability without changing its functionality. This frees up developers from having to manually optimise the code, saving valuable time. AI's ability to learn from vast datasets is enhancing its ability to generate more sophisticated code. As AI continues to be trained on increasingly diverse codebases, its ability to understand complex coding patterns and structures improves, allowing it to generate more contextually relevant and error-free code. However, AI's growing involvement in software development also raises questions about intellectual property and the ethics of using AI to create code. As AI systems become more involved in generating original code, the issue of who owns the rights to AI-generated software becomes increasingly important. The tech industry is still grappling with these questions, and the answers will likely evolve as AI tools become more integrated into the software development lifecycle.
Express Tribune
30-04-2025
- Business
- Express Tribune
AI now writes up to 30% of Microsoft's code: Microsoft CEO
A Microsoft logo is seen a day after Microsoft Corp's $26.2 billion purchase of LinkedIn Corp, in Los Angeles, California, U.S., June 14, 2016. PHOTO:REUTERS Microsoft CEO Satya Nadella revealed that artificial intelligence now generates up to 30% of the code across the company's software projects, highlighting a major shift in how modern software is developed. Speaking at Meta's LlamaCon AI developer conference alongside Meta CEO Mark Zuckerberg, Nadella said, 'I'd say maybe 20%, 30% of the code that is inside of our repos today and some of our projects are probably all written by software,' referring to AI systems. The two tech leaders discussed the growing role of AI in software development. Nadella noted that AI-assisted coding is gaining traction at Microsoft, with more success seen in languages like Python compared to more complex ones such as C++. The percentage of AI-generated code is expected to rise steadily. Zuckerberg did not specify Meta's own AI coding metrics but said the company is developing AI models that can autonomously build future versions of its Llama AI models. 'Our bet is that in the next year, maybe half the development is going to be done by AI,' he said. The comments come amid an industry-wide shift. Google CEO Sundar Pichai recently stated that over 30% of new code at Google is AI-generated. Microsoft CTO Kevin Scott has previously predicted that 95% of code could be written by AI by 2030. As tech giants automate more tasks, companies like Shopify and Duolingo have begun requiring proof that humans can outperform AI before justifying new hiring. The use of AI in software development marks a broader transformation in the global tech workforce and productivity strategies.
