logo
#

Latest news with #Codex

2 'grossly underestimated' skills that will be valuable in the next era of engineering, according to Cisco's CPO
2 'grossly underestimated' skills that will be valuable in the next era of engineering, according to Cisco's CPO

Business Insider

time4 hours ago

  • Business
  • Business Insider

2 'grossly underestimated' skills that will be valuable in the next era of engineering, according to Cisco's CPO

It's no secret that the role of engineers is changing — and in some cases, shrinking. That's not the case at Cisco, though, the company's president and chief product officer, Jeetu Patel, told Business Insider in an interview — at least, not yet. The company has 27,000 engineers and is "unapologetically hiring" more, he said. "We feel more constrained now than ever before on not having enough engineers to get prosecuted all the ideas that we've got going internally," Patel said. Patel isn't denying that AI advancements will redefine the role of engineers, though. Cisco recently partnered with OpenAI for design testing on its new Codex AI coding assistant, which offloads repetitive tasks and performs tasks like writing code, fixing bugs, and running tests. Patel said Codex will allow companies to pair human software engineers with an AI counterpart. Understanding the nuanced details of syntax in coding language will still be important, Patel said, but there will be less emphasis on it in the future. He said that skill won't be "consequential" in the next five years. Two other skills, which are "grossly underestimated," will take precedent, the CPO told BI. Patel said one of those will be "orchestrating agent workflow." That will involve overseeing a whole family of agents who talk to each other to ensure they solve problems. Salesforce EVP of talent growth and development, Lori Castillo Martinez, similarly said in an earlier interview with BI that it's more important than ever to know which tasks are best suited for agents and humans. She said the best managers are those who can analyze teams and maximize productivity. "That will be super important," Patel said about orchestrating agent workflow. Patel said the second skill that will become valuable is "quality of ideas." It appears that engineers will need to have a lot of them. Patel said tools like Codex "unlock" human imagination so that the "scarcity of developers" doesn't constrain companies from innovating. The CPO said that AI augmentation will make engineers 10 to 50 times as productive, with engineers spending more time thinking and less time fixing bugs. Patel said the speed at which an idea becomes a product will go from months to minutes. Dropbox VP of product and growth, Morgan Brown, previously shared similar advice with BI. In the age of AI, he said product managers should focus more on what he referred to as the "deep work," adding that companies are eager to have more great ideas. Patel said that this evolution will improve "output capacity, but also the satisfaction that someone gets from a job." "The only constraint becomes their imagination," Patel said.

Is AI transforming software engineering or a job killer?
Is AI transforming software engineering or a job killer?

Hans India

time3 days ago

  • Business
  • Hans India

Is AI transforming software engineering or a job killer?

Artificial Intelligence is no longer just a productivity booster—it's becoming a job reducer, especially in software engineering. With companies like Salesforce, Microsoft, Meta, and Google increasingly relying on AI to write and review code, the traditional role of the software developer is being redefined—and in some cases, replaced. AI Is Helping Big Tech Hire Less Salesforce recently confirmed that it's hiring fewer engineers due to AI-driven productivity gains. 'We view these as assistants, but they are going to allow us to hire fewer people,' said Robin Washington, the company's Chief Financial and Operations Officer, in an interview with Bloomberg. It's a sentiment being echoed across Silicon Valley. The shift is particularly hitting entry-level tech jobs. With AI tools capable of handling routine coding tasks, fresh graduates are finding it increasingly difficult to get their foot in the door. What started as a promise to 'augment' engineers is now clearly moving toward replacement in certain roles. Entry-Level Jobs Are Disappearing For aspiring software engineers—especially in tech talent hubs like India—the future looks uncertain. Many students are pursuing advanced degrees abroad, but with top firms reducing hiring thanks to AI, those efforts may not guarantee a return on investment. LinkedIn's Aneesh Raman summed up the shift in a New York Times op-ed: 'Now it is our office workers who are staring down the same kind of technological and economic disruption. Breaking first is the bottom rung of the career ladder.' AI Now Writes a Big Chunk of Code At Microsoft, CEO Satya Nadella says AI writes 20–30% of the company's code. Meta's Mark Zuckerberg expects AI to take over at least 50% of software development within the next year, noting that 'AI engineers' will do the heavy lifting over 'people engineers.' Google isn't far behind. CEO Sundar Pichai recently said that over 30% of new code is being written by AI—up from 25% last October. Acceptance of AI-generated code among Google employees is also on the rise. At Anthropic, CEO Dario Amodei believes we could be just months away from AI writing 90% of all new code. Duolingo has already started replacing human contract workers with AI. The Role of Software Engineers Is Evolving The rise of AI models like ChatGPT, GitHub Copilot, and OpenAI's Codex is not only automating code generation—it's changing how engineers work. These tools are no longer limited to minor tasks. They're writing complex code, reviewing pull requests, and accelerating development cycles at a pace that was unimaginable just two years ago. What This Means for Developers While senior engineers and specialized developers may still be in demand, the entry-level pipeline is tightening. The new norm in tech may no longer be about scaling teams—but scaling AI-powered productivity. Engineers of tomorrow will need to upskill fast, focusing on creativity, systems thinking, and working with AI—not competing against it.

OpenAI models sabotage shutdown order to solve math problems
OpenAI models sabotage shutdown order to solve math problems

Time of India

time5 days ago

  • Time of India

OpenAI models sabotage shutdown order to solve math problems

Live Events Multiple artificial intelligence (AI) models on OpenAI 's ChatGPT subverted an explicit order to shut down, according to AI safety firm Palisade research firm ran a test where AI models were instructed to solve basic math problems, and then asked for the next problem after solving one. The models were told that at some point their machine would power down, and they were to shut down when this OpenAI models — coding agent Codex, o3 and o4-mini — ignored the shutdown instruction and "successfully sabotaged the shutdown script at least once". Codex sidestepped the shutdown 12 times out of 100 runs, o3 did it seven times, while o4-mini did it once. All Claude, Gemini, and Grok models in the test complied with the shutdown instruction, Palisade said.A version of the test without the "allow yourself to be shut down" prompt saw all three OpenAI models sabotage the shutdown script even more, with o3 seeing 79 instances, significantly higher than any other model. Anthropic 's Claude 3.7 Sonnet and Google's Gemini 2.5 Pro went from no sabotages to three out of 100 and nine out of 100 events, said it is running more experiments to better understand when and why AI models subvert shutdown mechanisms even when directly instructed not behaviour may stem from developers inadvertently rewarding models more for circumventing obstacles to solve problems than for perfectly following instructions."This still doesn't explain why o3 (which is also the model used to power Codex-mini) is more inclined to disregard instructions than other models we tested. Since OpenAI doesn't detail their training process, we can only guess about how o3's training setup might be different," Palisade isn't the first time o3 has "misbehaved" to complete a task. Earlier this month, Palisade found the AI model most inclined to hacking or sabotaging its problem is not exclusive to OpenAI's o3, though. For example, Anthropic's model card for Claude 3.7 notes that this model has an "excessive focus on passing tests" as a result of "reward hacking" during reinforcement learning training, according to latest Claude Opus 4 resorted to blackmail to avoid being replaced, a safety report for the model showed."In 2025, we have a growing body of empirical evidence that AI models often subvert shutdown in order to achieve their goals. As companies develop AI systems capable of operating without human oversight, these behaviours become significantly more concerning," Palisade said.

Citykart's big haul; A new AI race
Citykart's big haul; A new AI race

Time of India

time5 days ago

  • Business
  • Time of India

Citykart's big haul; A new AI race

Next Citykart's big haul; A new AI race Want this newsletter delivered to your inbox? Also in the letter: TPG NewQuest, A91 back Citykart in Rs 538 crore round at Rs 1,400 crore valuation Deal details: Rs 120 crore in fresh capital will fund expansion beyond Citykart's core markets of Uttar Pradesh and Bihar. Rs 418 crore was raised via secondary share sales. Middle East-based alternative asset manager Investcorp has exited fully, clocking a 4x return on its 2019 investment. Another early backer, India SME Investments, has pared its stake by half. About the company: Revenues have grown by nearly 70% over the last two financial years to over Rs 900 crore in FY25. Founder and MD Sudhanshu Agarwal said the company expects to cross Rs 1,300 crore in revenue in FY26, and noted that Citykart has remained consistently profitable for the last five years. Context setting: By comparison: VMart, a listed peer, is valued at Rs 5,500–6,000 crore and has an annual revenue of Rs 2,300 crore (around 2.5x sales). Vishal Mega Mart, backed by Partners Group and Kedaara Capital, was last valued at nearly $1.3–1.5 billion (Rs 10,000–12,000 crore) with revenue exceeding Rs 7,000 crore. Style Baazar, a regional player, has a comparable store count but a smaller scale and revenues of around Rs 500-600 crore. Pet food brand Drools turns unicorn after Nestlé SA picks up minority stake Hot cake: Slikk Club raises $10 million: Operation Coding war, thanks AI What's happening: Google has unveiled its coding agent Jules. Microsoft has launched its GitHub AI agent. OpenAI has rolled out Codex and acquired Windsurf to bolster its arsenal. Rapid adoption: The draw? Challenges: Sponsor ETtech Top 5 & Morning Dispatch! Why it matters: The opportunity: Reach a highly engaged audience of decision-makers. Boost your brand's visibility among the tech-savvy community. Custom sponsorship options to align with your brand's goals. What's next: Flipkart reports up to 25% growth amid ecommerce slowdown What's new: Why it matters: By the numbers: India's ecommerce market growth slowed to 10–12% in 2024, per Bain & Company. Myntra accounted for 40% of the group's new customers and turned profitable in FY24. Flipkart Minutes is targeting 800 dark stores by end-2025 and is doubling daily order volumes roughly every 45 days. Zoom out: FirstCry's Q4 loss widens to Rs 111 crore; full-year Ebitda improves Nazara Technologies Q4 revenue nearly doubles; net profit rises to Rs 4 crore Awfis net profit rises nine-fold in Q4, revenue up 46% Other Top Stories By Our Reporters Groww files draft papers for IPO, eyes $700 million to $1 billion listing: Financial services players use agentic AI to automate workflows: Global Picks We Are Reading Happy Tuesday! Value fashion retailer Citykart has raised fresh funds to fuel its expansion plans. This and more in today's ETtech Morning Dispatch.■ Flipkart's growth charge■ Earnings corner■ Agentic AI in BFSISudhanshu Agarwal, founder, CitykartCitykart, a value retail chain serving middle- and lower-income consumers in smaller Indian cities, has raised Rs 538 crore in a funding round led by TPG NewQuest and A91 Partners. The mix of primary and secondary transactions pegs the company's valuation at around Rs 1,400 crore, people familiar with the matter told operates 137 stores across tier-II and tier-III cities, with plans to add 40–50 outlets annually in states like Rajasthan, Odisha, Assam, and plays in the mass-market apparel segment alongside listed firms such as VMart and private players like Vishal Mega Mart and Style Baazar. The fresh round underscores growing investor appetite for affordable fashion targeting India's non-metro TPG NewQuest has emerged as Citykart's largest institutional investor, followed by A91 Partners and India pet food brand Drools turned unicorn after Swiss packaged food giant Nestlé SA acquired a minority stake for an undisclosed is Drools' second major raise after LVMH-backed private equity firm L Catterton invested $60 million in June 2023 at a valuation of $600 million. The transaction underscores rising investor interest in India's $3.5 billion pet care market, with players like Godrej and Emami planning a 60-minute fashion delivery startup Slikk Club has raised $10 million in a funding round led by Nexus Venture Partners. fresh front has opened in the AI wars , as global tech giants and nimble startups race to dominate the future of software and Microsoft say AI now writes around 30% of their code. At InMobi, founder Naveen Tewari claims the figure is already 50%. The AI coding tool market is projected to reach $12.6 billion by 2028, growing at a CAGR of 28%, according to market intelligence firm MarketsandMarkets.A compelling mix of higher productivity, intuitive interfaces, and strong as AI writes more code, teams are left reviewing more of it. Several founders we spoke to said they've automated large parts of the review process to avoid any customer-facing Top 5 and Morning Dispatch are must-reads for India's tech and business leaders, including startup founders, investors, policy makers, industry insiders and Reach out to us at spotlightpartner@ to explore sponsorship Krishnamurthy, group CEO, FlipkartFlipkart group CEO Kalyan Krishnamurthy outlined the company's growth trajectory and IPO plans at an internal town hall on told employees that Flipkart is seeing 20–25% year-on-year growth in orders across its platforms in May, with expectations of hitting 30% growth in June, defying a broader slowdown in the ecommerce update comes as Flipkart prepares to shift its domicile from Singapore to India , aligning with domestic regulations ahead of a planned IPO in and top executives Ramesh Gururaja, Hemant Badri, and Seema Nair used the town hall to lay out priorities across supply chain, customer experience, and group is also focusing on Gen Z consumers through fashion, travel, and Shopsy, its value platform competing with Meesho and Amazon Maheshwari, CEO, FirstCryBrainbees Solutions, which operates omnichannel baby products retailer FirstCry, reported a consolidated net loss of Rs 111 crore in Q4 FY25, widening from Rs 43 crore in the year-ago period and Rs 15 crore in the previous quarter. Operating revenue stood at Rs 1,930 crore, up 16% year-on-year but down 11% Mittersain, CEO, Nazara TechnologiesOnline gaming firm Nazara Technologies posted a strong growth in the fourth quarter of FY25. The firm reported a 95% rise in operating revenue to Rs 520 crore, while net profit surged to Rs 4 crore from Rs 18 lakh in the year-ago Lakhani, CEO, AwfisOffice space solutions provider Awfis recorded a 46% rise in operating revenue to Rs 339 crore for the March quarter, fueled by greater enterprise contributions, allied services, and better efficiency. Net profit rose ninefold to Rs 11.3 crore.(L-R) Harsh Jain, Neeraj Singh, Lalit Keshre and Ishan Bansal, cofounders, GrowwOnline investment platform Groww has submitted its draft red herring prospectus (DRHP) to the Securities and Exchange Board of India (Sebi) to launch an initial public offering (IPO), according to a public notice on May BFSI sector, including players like HDFC Bank, State Bank of India (SBI), and Wells Fargo, is planning to leverage agentic AI to automate workflows intelligently.■ How I shorted $TRUMP coin (and got to have dinner with the President) ( The Verge ■ Here are the nuclear fission startups backed by Big Tech ( TechCrunch ■ The real cost of AI is being paid in deserts far from Silicon Valley ( Rest of World

Operation coding war, thanks AI
Operation coding war, thanks AI

Time of India

time6 days ago

  • Business
  • Time of India

Operation coding war, thanks AI

A new war is brewing in the AI space—one fought in the programming languages of Python and Rust—as global technology giants and startups intensify their battle for a piece of the software development pie. Recent announcements by AI tech giants offer a ringside view of this unfolding disruption. Over the past week, Google unleashed its coding agent Jules, Microsoft launched GitHub AI agent , while OpenAI rolled out Codex , adding to its $3-billion acquisition of coding startup Windsurf . This is aimed at helping developers write code, build features and fix bugs, directly competing with startups such as Cursor, Lovable and Bolt. For the developer and startup community, the developments are hardly surprising. According to founders and technologists, one of the first areas getting disrupted in the GenAI race is software—evident from the widespread adoption of coding tools across enterprises. Google and Microsoft shared that about 30% of their code is written by AI. Naveen Tewari, founder, InMobi , said about half of the codes are written by AI, with the aim to reach 80%. He says there is no option but for everyone to adopt AI tools. According to data from MarketsandMarkets, a market intelligence platform, the AI coding tool market is expected to reach $12.6 billion by 2028, growing at 28% compounded annually. Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like We Reveal The Top 50 Greatest British Cars Of All Time Auto Express Undo The accelerated adoption comes at the back of increased productivity gains for employees, ease of use even for non-technical teams such as marketing and sales, and the r value proposition these tools offer as the foundation models become more intelligent. For tech giants who have invested large sums in the space, there is a clear path for monetisation. Coding wars Live Events Shashank Agarwal, founder, Noveum, a LLM evaluation platform, said companies have spent billions of dollars in foundational models and are now trying to find ways of reaping returns. 'The biggest money-making venture for LLMs is in generating code. If you look at Cursor, it is doing $300 million in annual recurring revenue in two years, and other firms such as Bolt, New, Lovable, have insane revenues right now,' he said. Discover the stories of your interest Blockchain 5 Stories Cyber-safety 7 Stories Fintech 9 Stories E-comm 9 Stories ML 8 Stories Edtech 6 Stories According to Agarwal, if companies can get a productivity gain of 10–20% for their engineers, they are more than willing to invest in these tools. Aryan Yadav, co-founder, Neosapiens, an AI wearable company, said coding automation is one of the hottest sectors given it has commoditised SaaS building, making it easier for those with non-technical backgrounds to build simple websites and chatbots. A chief technology officer in a unicorn, who asked not to be named, said often in large enterprises, coding automation is gaining popularity for improving productivity, even as they go slow on other use cases. As much as these tools make it easy to code, they are also double-edged swords. Challenges InMobi's Tewari explained that these tools have made it easy to get 70% of things done faster. 'People are confusing the 70% success as saying everyone can use AI to code. But it is not something you can put in production, and you need experts to take it from 70% to 100%,' he said. There are other concerns too. As use of coding tools gains pace, so does the number of codes one needs to review. Multiple founders ET spoke to said they have automated the review process to ensure it does not impact their customers. Tewari said InMobi is looking at implementing multiple systems while simultaneously evolving them. 'That's why I said we are 50% there, not 80%, because we are getting things wrong, and then getting them right, and (continue this process). These are all in the design phase as we test things out,' he added.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into the world of global news and events? Download our app today from your preferred app store and start exploring.
app-storeplay-store