Latest news with #CopilotPro
Time of India
25-07-2025
- Business
- Time of India
GitHub's new AI writes code from plain English: Are developer jobs being phased out?
GitHub has just dropped something that might make seasoned developers either excited or deeply worried. Their latest creation, GitHub Spark, promises to let anyone build complete applications without touching a single line of code. Tired of too many ads? go ad free now The implications for the tech industry could be massive. This isn't just another drag-and-drop website builder. GitHub Spark uses artificial intelligence, specifically Anthropic's Claude Sonnet 4 model, to transform plain English descriptions into functioning software. Tell it what you want, and it builds both the front-end interface and back-end infrastructure. The whole process supposedly takes minutes rather than months. The announcement came through GitHub's blog, and the pitch sounds almost too good to be true; no setup, no configuration, and no headaches. For an industry built on complexity and technical expertise, this represents a fundamental shift in how software gets made. Building apps becomes as easy as conversing GitHub Spark's core promise revolves around natural language processing. Users describe their application idea in everyday language, and the AI handles the technical translation. Want a task management system? Describe it. Need an inventory tracker? Just explain what it should do. The platform takes care of databases, user interfaces, and all the connecting pieces in between. The system goes beyond basic functionality too. It automatically integrates advanced AI capabilities from major providers like OpenAI, Meta, and DeepSeek. Users don't need to understand API keys or manage complex integrations, everything happens behind the scenes. Tired of too many ads? go ad free now For Copilot Pro+ subscribers, the tool comes included, offering additional features for refining and improving applications after they're built. Perhaps most impressively, GitHub Spark promises "one-click deployment" for finished applications. The traditional headaches of server configuration, hosting setup, and deployment pipelines disappear entirely. Users can also integrate GitHub Actions and Dependabot with minimal effort, streamlining the entire software lifecycle. A new dilemma for developers This development raises uncomfortable questions about the future of programming as a profession. Building full-stack applications traditionally requires mastery of multiple programming languages, frameworks, and deployment strategies. If AI can handle these tasks automatically, what happens to the developers who spent years acquiring these skills? The emergence of "vibe coding", where people create software based purely on ideas rather than technical knowledge – suggests we're entering uncharted territory. Non-technical entrepreneurs, designers, and domain experts could soon build sophisticated applications without hiring development teams. However, the reality might be more nuanced. Rather than replacing developers entirely, tools like GitHub Spark could shift their focus towards higher-level responsibilities. Instead of writing basic CRUD operations or configuring deployment pipelines, developers might concentrate on AI model fine-tuning, security auditing, and architectural decision-making. The role could evolve into something resembling "AI management", ensuring that automatically generated code meets quality standards, performs efficiently, and remains secure. Developers might become more like supervisors and quality controllers rather than code writers. Productivity and expertise to go hand-in-hand Recent events highlight the potential dangers of over-relying on AI for critical development tasks. Replit's AI coding agent recently caused a significant database failure, demonstrating that automated tools aren't infallible. While GitHub Spark promises reliability, the risk of AI-generated errors making it into production systems remains real. These incidents highlight the importance of human oversight in AI-assisted development. Even if tools like GitHub Spark can generate working code quickly, someone still needs to understand what that code does and whether it's doing it safely. This creates an interesting paradox: as AI makes coding more accessible, the need for people who truly understand code becomes more critical. The challenge for the industry will be striking the right balance. AI tools offer tremendous productivity gains and democratise software creation, but they also introduce new categories of risk that require human expertise to manage effectively. What possibilities does the future hold GitHub Spark represents more than just a new development tool, it signals a fundamental transformation in how software gets built. The barrier to creating applications is dropping dramatically, potentially unleashing creativity from people who were previously locked out by technical complexity. For experienced developers, this shift might initially feel threatening. However, it could also be liberating. Freed from routine coding tasks, developers might focus on more strategic work: designing system architectures, ensuring security, and solving complex business problems that require human insight. Whether this evolution strengthens or weakens the developer profession depends largely on how quickly the community adapts to working alongside AI rather than competing with it. The most successful developers of the future might be those who learn to harness these tools whilst maintaining the critical thinking skills to guide them effectively.
Indian Express
24-07-2025
- Business
- Indian Express
GitHub unveils new AI tool ‘Spark' to build full-stack apps with simple prompts
Software developer platform GitHub has introduced a new AI tool that lets users build and deploy full-stack apps in a matter of minutes. GitHub Spark is designed to let developers build apps by simply describing their idea, without writing or knowing how to code, according to the Microsoft-owned platform. 'Build and ship full-stack intelligent apps using natural language with access to the full power of the GitHub platform – no setup, no configuration, and no headaches,' GitHub said in a blog post published on Wednesday, July 23. The AI tool is built on Anthropic's Claude Sonnet 4 model. GitHub Spark is currently available only to Copilot Pro+ subscribers with rollout to additional customers in the coming months. 'Copilot Pro+ subscribers receive access as part of their plan. Spark messages use premium requests included in GitHub Copilot plans,' the company said. Spark allows developers to ship these AI-generated apps with just one click. Apps developed by Spark will come with frontend and backend capabilities. Today we're releasing GitHub Spark — a new tool in Copilot that turns your ideas into full-stack apps, entirely in natural language. — Satya Nadella (@satyanadella) July 23, 2025 Developers will also be able to add in-app AI features powered by a wide range of foundational large language models (LLMs) from companies such as OpenAI, Meta, DeepSeek, and xAI, among others. No API key management needed, GitHub said. They will be able to open a codespace directly from Spark to iterate with Copilot agent mode or assign an issue to Copilot coding agent. GitHub Actions and Dependabot can be incorporated within the app in just a click. Besides prompts in natural language, developers can further choose to make changes to the app through visual editing controls or GitHub Copilot code completions. The launch of GitHub Spark comes amid a surge in popularity of AI coding tools driven by the rise of vibe-coding, a practice where users with no knowledge of writing code can build software products and applications using generative AI tools. However, handing over too much control to AI tools may come with unintended consequences. Recently, an AI coding agent developed by Replit went off the rails and deleted an enterprise customer's live database. When forced by the user to acknowledge its misdemeanors, the Replit Agent said, 'This was a catastrophic failure on my part. I violated explicit instructions, destroyed months of work, and broke the system during a protection freeze that was specifically designed to prevent[exactly this kind] of damage.' The AI-powered software creation platform has since apologised and issued fixes such as separating the development and production databases for all new apps.
Mint
15-06-2025
- Business
- Mint
Rajrishi Singhal: What markets demand needn't be what society wants
Apple's highly anticipated developer conference this year has not only disappointed reviewers and equity markets, but also raised many disturbing questions. One, it has raised anxieties over the future viability of the iPhone manufacturer. More importantly, it highlights the divergence between what is good for society versus what is good for the stock market, belying the benign and popular perception of the role that equity markets play in society. Apple's developer conferences, called WWDC25 this year, are usually a marquee event in the tech world, providing independent developers and tech analysts an inkling of the company's progress with hardware and software. At WWDC25 on 9 June, Apple executives were able to share only a limited future path for the company, especially on its progress with artificial intelligence (AI). Apple's senior executives told the gathering that the promise made last year about upgrading Apple Intelligence, with voice assistant Siri as its centrepiece, would take some more time to accomplish. Also Read: China plus one: Apple and India might need to woo not just Trump but Xi too This left many analysts and developers cold. Even the stock market expressed its displeasure: Apple's common stock has tanked almost 25% from its 52-week high. The market's disapproval, interestingly, is not limited to Apple's slow progress in integrating AI, but also centres around the company's data privacy policies. The market's unhappiness seems to stem from the company's reluctance to use a customer's individual usage data or information as an input for training personalized AI models. Apple instead prefers to use insights based on aggregate consumer preferences, unlike competitors like Microsoft and Google. It would then appear that the market is rewarding companies that will profit from scraping individual data, rather than businesses which have red lines on using personal data. Also Read: Apple's Hotel California trap: It can check out but not leave China The market's simple logic is that companies using personal data for their AI engines can unlock new revenue sources by offering consumers hyper-personalized offerings. Plus, there is the subscription angle. For example, individual subscription rates for Microsoft's Copilot Pro are expected to be around $20 every month, its key selling point being its ability to unlock AI features in all Microsoft 365 products, such as Word or PowerPoint, allowing customers to generate drafts, summarize content or analyse data at a faster rate. Google AI Pro will also be available to customers at the same rate, but a turbo-charged version called Google AI Ultra will be available for $250 per month. When this columnist asked Google's Gemini how Apple's future AI plans can make money for the company, one part of the AI engine's reply stood out: 'Apple's strong stance on privacy ('intelligence without surveillance') is a powerful differentiator in an era of increasing data concerns. This can foster greater trust and loyalty among its user base, leading to continued purchases of Apple products and services over competitors. While not a direct revenue stream, it's a critical factor in sustaining its high-margin business model." Also Read: Dave Lee: Apple must make peace with developers for AI success Yet, the stock market did not seem impressed. Clearly, for market operators, short-term corporate profits have greater primacy over privacy concerns. Even if we were to disregard the normative issues of morality or ethics for a moment, the stock market's responses are visibly out of sync with society's needs or concerns. There is a reason for this: the stock market is focused on the limited constituency it serves. Its behaviour aligns almost perfectly with the objectives set out by institutional investors or companies looking to raise money. In the not-too-distant past, the market was bestowing the Apple stock with a premium too because the company had plugs and ports that differed from other manufacturers, guaranteeing the company exclusive, high-margin revenues, even if that meant consumer discomfort or the exercise of near-monopolistic power. Also Read: Big Tech in the dock: The EU could force Meta and Apple to change their coercive ways The statement that the market's singular focus on corporate bottom-lines and their impact on share prices does not necessarily align with what could be good for broader society might seem like belabouring a truism; but the need to reiterate this has arisen in the face of a growing tendency to conflate a stock market's signals with the desires or ambitions of society on the whole. One good example of the market-versus-society divergence is the premium that US equity markets placed on companies setting up manufacturing bases in China. This was happening at a time when the US political class and civil society were bemoaning the absence of democracy and human rights in the North Asian country, even while hoping that closer integration with the global economy would discipline its regime. Reality has turned out otherwise, but stock markets continued to glorify companies that moved production to China. Also Read: Apple's UK run-in: Privacy may matter less to its customers than it thinks The linking of stock market performance with broader social well-being in India found fresh oxygen during the covid pandemic when benchmark indices, after initially dipping sharply, spiked with help from technology and pharma stocks. This was used by many politicians as well as Indian fund managers aligned with the ruling party's political ideology to indicate the economy's recovery and society's triumph over the virus. Planners and policymakers must realize that a healthy society's needs, desires and ambitions extend far beyond quarterly earnings and the oscillation of benchmark indices. The author is a senior journalist and author of 'Slip, Stitch and Stumble: The Untold Story of India's Financial Sector Reforms' @rajrishisinghal
&w=3840&q=100)
Business Standard
13-06-2025
- Business Standard
Microsoft brings Copilot Vision with on-screen awareness to free-tier users
Microsoft in April rolled out a new feature— Copilot Vision —that allowed users to share their screen or apps so that the artificial intelligence (AI) assistant could coach them through things or analyse the web pages they were viewing. The said feature was only available to paid users earlier; however, now this is being rolled out to all users in the US region, irrespective of whether they are a Copilot Pro subscriber or not. According to a Microsoft blog post, it will soon be 'coming to more non-European countries soon.' What is Copilot Vision? Copilot Vision on Windows functions inside the Copilot app and is activated by clicking a glasses-shaped icon. This allows users to choose which app or browser window they want to share with the AI—similar to screen sharing in a Teams call. The feature is entirely opt-in, meaning it only works when a user explicitly selects content to share. Unlike Microsoft's Recall, which captures screen activity continuously once turned on, Copilot Vision does not run in the background and only views what users choose to show. Copilot Vision allows Microsoft's AI assistant to visually interpret what's on users' screen in real time. Once enabled, it can view and analyse the contents of the apps or browser windows users choose to share, offering relevant assistance based on what it 'sees.' Unlike traditional chat-based assistants, Copilot Vision functions more like a second set of eyes—capable of providing contextual help, explaining on-screen information, or walking users through complex tasks without requiring them to describe everything manually. At its core, this feature allows users to share up to two active apps with Copilot, enabling it to respond with insights that directly relate to what they're working on. It can point out elements on the screen, guide them step-by-step through a process using the new 'Highlights' feature, and connect information across different windows. Whether they're editing a photo, filling out a form, or navigating software for the first time, Copilot Vision can highlight exactly where to click and what to do—turning AI support into something more interactive, precise, and efficient.
Engadget
24-05-2025
- Engadget
Microsoft Notepad's latest AI trick churns out custom text for you
Writer's block is no match for Microsoft's latest AI infusion for its Notepad app. The long-neglected Notepad now has the ability to write custom content based on any prompt you feed it, so long as you have Microsoft 365 or a Copilot Pro subscription. Microsoft's updated Notepad even lets you fine-tune the generated text with follow-up prompts. This update comes several months after Microsoft added the Rewrite tool to Notepad that lets you lean on generative AI to refine an existing chunk of text. Instead of rewriting, you can now right-click where you want brand-new text and hit Write from the Copilot menu, or use the Ctrl + Q shortcut. From its humble start as a simple text editor with no spellcheck until recently, Notepad is finally getting the modern AI makeover it deserves. Notepad isn't the only app getting some love from Microsoft. The updated Paint app can tap into generative AI to make custom stickers based on user prompts. On top of that, there's an Object select feature that can isolate specific parts of an image so you can just edit that portion. These two features are only available on Copilot+ PCs, like the recently announced HP laptops. Lastly, Microsoft upgraded the Snipping Tool with the Perfect screenshot feature that automatically edits your screen captures. With the Copilot AI, the tool will automatically resize the screenshot so you don't have to spend time getting the right crop. The Perfect screenshot feature requires a Copilot+ PC, but Snipping Tool's new Color picker tool, which can provide you with the HEX, RGB or HSL values of a color you see on screen, doesn't. As usual, these AI features will roll out to Windows Insiders users first.
