AI doing up to 50% of work at Salesforce: CEO Marc Benioff
Reuters Salesforce CEO Marc Benioff, in an interview with Bloomberg, said that the company is using artificial intelligence (AI) to do 30% to 50% of the work.Similar statements have been made by other tech leaders, including Microsoft CEO Satya Nadella and Google chief Sundar Pichai, who have said that AI does a significant portion of work like coding in their companies.
Benioff continued, 'All of us have to get our heads around this idea that AI can do things that before, you know, we were doing, and we can move on to do higher-value work.' When the interviewer Emily Chang asked if AI could replace him one day, Benioff joked, 'I hope so. I mean, of course, I'm partially kidding. You know that, but we're becoming more automated.'According to Bloomberg, Salesforce has already let go of over 1,000 employees this year. While some may be moved into new roles, large-scale job cuts like these are becoming more common with the growing adoption of AI.Last month, Salesforce acknowledged that its adoption of AI tools has enabled the company to scale back on recruitment. During an analyst call, chief financial and operations officer Robin Washington said the company is reducing hiring, with 500 customer service staff to be moved into other roles, an adjustment expected to save $50 million.A few months ago, Benioff went so far as to say that the company wouldn't be hiring any more engineers this year due to significant productivity gains from AI.However, it is worth noting that despite the slowdown in technical hiring, the company is expanding its sales team.As of January 31, Salesforce had a workforce of around 76,500. Elevate your knowledge and leadership skills at a cost cheaper than your daily tea. The bike taxi dreams of Rapido, Uber, and Ola just got a jolt. But they're winning public favour
Second only to L&T, but controversies may weaken this infra powerhouse's growth story
Punit Goenka reloads Zee with Bullet and OTT focus. Can he beat mighty rivals?
3 critical hurdles in India's quest for rare earth independence
HDB Financial may be cheaper than Bajaj Fin, but what about returns?
Why Sebi must give up veto power over market infra institutions
These large- and mid-cap stocks can give more than 23% return in 1 year, according to analysts
Are short-term headwinds from China an opportunity? 8 auto stocks: Time to be contrarian?
Buy, Sell or Hold: Motilal Oswal initiates coverage on Supreme Industries; UBS initiates coverage on PNB Housing
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
5 hours ago
- Time of India
How there has been a shift in companies that use H-1B visas in the US, they are no longer technology companies
Representative Image Silicon Valley's drive for innovation has long relied on the H-1B visa program to attract top-tier global talent in science and engineering. However, new data obtained by Bloomberg News reveals that a wider range of industries, including banks and telecommunications companies, are among the largest users of the programme -- often through staffing and outsourcing companies that secure nearly half of the 85,000 new H-1B visas issued annually. These companies, acting as visa middlemen, are reshaping the programme's purpose, raising questions about its impact on wages and U.S. workers. According to Bloomberg's analysis, covering new H-1B hires from May 2020 to May 2024, Citigroup Inc. added 3,000 H-1B workers, outpacing tech giants like Nvidia, Oracle, and Qualcomm. However, unlike the high-skill researchers and engineers typically associated with tech companies, about two-thirds of Citi's H-1B workers were IT contractors sourced through staffing and outsourcing agencies. These workers, often reportedly paid significantly less than direct hires, highlight a growing trend where non-tech companies leverage the visa program for lower-cost labor. 'This is the tip of the iceberg,' said Susan Houseman, a senior economist at the W.E. Upjohn Institute for Employment Research, in an interview with Bloomberg. 'While there's been a national debate about the U.S. reliance on imported goods, not enough attention has been paid to the offshoring of service jobs -- not because it doesn't happen or isn't important, but because we don't have good data on it.' Lottery fraud in H-1B visas by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Sharp Design, Smoother Drives. Toyota Glanza Learn More Undo The H-1B visa, established by Congress in 1990 to bolster U.S. competitiveness in emerging tech industries, has become oversubscribed, leading to annual lotteries for the limited visas available. Staffing and outsourcing firms, functioning as visa middlemen, have exploited this system. These firms provide corporations with lower-level IT workers or facilitate the offshoring of back-office functions. Bloomberg's data, obtained via a Freedom of Information Act lawsuit against the Department of Homeland Security, identifies major U.S. companies like Capital One Financial Corp., Verizon Communications Inc., AT&T Inc., and Walmart Inc. as heavy users of these visa middlemen. Until recently, some middlemen manipulated the visa lottery through a practice known as 'multiple registration,' where they submitted numerous applications for the same worker to increase their odds. The U.S. Citizenship and Immigration Services (USCIS) labeled this practice fraudulent in a 2023 report and implemented rule changes last year to curb it. Bloomberg's investigation also identified 13 staffing firms flagged by USCIS for such tactics, with at least six supplying workers to Capital One. Over half of Capital One's 905 H-1B contract workers during the four-year period were linked to multiple registrations, the highest proportion among the top 10 companies analyzed. Capital One relied on 429 staffing firms, 361 of which used multiple registrations. In response, a Capital One spokesperson told Bloomberg that the company was unaware of any government accusations of visa fraud by its vendors but would 'take appropriate action' if such issues arose. Verizon and Capital One emphasized that they require suppliers to comply with applicable laws, while AT&T, Walmart, and USAA declined to comment. What the shift in H-1B visas mean for salaries The data also reveals significant pay disparities. H-1B contractors are often paid far less than direct hires, even for similar roles. Bloomberg's analysis of the top 10 end-clients shows that among nearly 5,300 H-1B 'software developers' hired from 2020 to 2024, over 75% were contractors, earning roughly $48,000 less on average than direct hires, even when controlling for education and age. One in three contractors received the minimum salary required by the Department of Labor. Steve Hall, Chief AI Officer at Information Services Group Inc., told Bloomberg that part of the pay gap reflects contractors performing less technical roles, such as liaising between U.S. clients and offshore teams. However, critics argue that the reliance on visa middlemen distorts the H-1B program's original intent. Labor advocates contend that it undercuts U.S. workers, creates a vulnerable workforce with fewer protections, and tilts the labor market in favor of employers. The lack of comprehensive data on pre-existing H-1B contractors and offshored jobs complicates efforts to assess the program's full impact. As Houseman noted, the offshoring of service jobs remains understudied, yet its implications for the U.S. labor market are significant. With companies across industries increasingly turning to visa middlemen, the H-1B program's role in shaping the workforce continues to spark debate. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Time of India
5 hours ago
- Time of India
Master prompt engineering the fast way - what Google's 9-hour course teaches you in 10 minutes
Prompt engineering is the flavour of the era, and Google has now come up with a 9-hour course to master the modern art. AI is everywhere now — it helps us write, create images, answer questions, and work faster. ChatGPT made AI super popular and now many AI tools are being used all over the world. All these tools work based on how we talk to them — that's called prompting. Prompting means giving instructions to the AI in a way that it understands and gives the output we want, as per reports. Google course structure Start writing prompts like a pro. This teaches you how to write basic and clear prompts to get better answers. Design prompts for everyday work tasks. Helps you learn how to use prompts for your daily office work or school tasks. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like One of the Most Successful Investors of All Time, Warren Buffett, Recommends: 5 Books for Turning... Blinkist: Warren Buffett's Reading List Click Here Undo Use AI for data analysis and presentations. Shows how to ask AI to help you understand data and make slides or reports. Use AI as a creative or expert partner. Trains you to use AI as a teammate who can help you think creatively or give expert-like advice, as stated by The Indian Express report. Here are them Modules breakdown as mentioned by Medium: Live Events Module 1: How to write prompts like a pro This module teaches you how to talk to AI properly so it gives better answers. Google shares a simple 5-step method called T.C.R.E.I.: Task – Say exactly what you want. Context – Give background info to help AI understand better. References – Show examples to guide the AI. Evaluate – Check if the answer is good. Iterate – If it's not perfect, rewrite your prompt and try again. Fun memory trick: 'Tiny Crabs Ride Enormous Iguanas.' You also learn to break big prompts into short ones. Try different wording. Add limits or rules to get more focused replies. ALSO READ: 2 per 20: The viral blood sugar hack that doesn't involve exercise or diet overhauls Module 2: Use AI for daily work stuff This part teaches how AI can help with daily office or school work. Some things you can do: Write better emails (faster and clearer). Create content like blog posts, newsletters, or social media ideas. Brainstorm new ideas (like campaign slogans or product names). Summarize long documents into short bullet points. Just use the same 5-step method from Module 1 to get better results. Module 3: Use AI for data and presentations This module shows how AI can help with numbers and slides. For data analysis: AI can help read spreadsheets, find patterns, or do simple math. Example: 'Find the average sales per customer in this Google Sheet.' For presentations: AI can write outlines or make slides based on your points. Example: 'Create slides for our sales report with 3 main sections: numbers, wins, and goals.' Module 4: Use AI as a creative or expert partner This part is about using AI for big ideas or expert help. You learn special techniques: Prompt chaining – Ask AI questions step by step to build a big answer. Chain of thought – Ask AI to explain its thinking. Tree of thought – Ask AI to give multiple ideas/solutions. You also learn how to create AI agents — smart bots that act like a teacher, coach, or role-player. Example: A mock interview bot to practice job questions. Plus, there's meta-prompting — where you ask AI to help you write better prompts. FAQs Q1. What is prompt engineering in AI? Prompt engineering means giving clear instructions to AI so it understands what you want and gives better results. Q2. Why is prompt engineering important? Because AI tools work based on how we ask them questions. Good prompts = good answers.
Hindustan Times
8 hours ago
- Hindustan Times
Clearing the fog on the state of India-US relations
'The administration is bullish on India' is how a senior US official put it to me last week in Washington D.C. This sentiment would seem at odds with the broader reporting on the US-India relationship. In a Financial Times newsletter on India, one writer argued that the Indian Prime Minister (PM) 'made the mistake of counting on his warm personal connection with Trump'. The general assertion being that the Indian government has mortgaged this crucial relationship to 'personal friendships' alone. Others suggest that the US President's recent luncheon with Asim Munir, the recently decorated Pakistani Field Marshal, and his 'sneaky attempt' to bring PM Narendra Modi and Munir into the same room in the White House is 'threatening the future of US-India partnership'. Structures like TRUST were created for top political leaders to monitor progress on crucial initiatives. (REUTERS) Between social media and popular reporting, it would seem as though this relationship has been iced. Yet, in meetings with over 30 officials, experts, think tankers, and industry representatives last week, the story that emerged was diametrically opposed to the one that has been paraphrased above. Modi's engagements with Trump matter more than it is perhaps realised. It clearly provides a political basis of what can be achieved between the two countries, even at this time of shrinking administrative capacities in the US, and the many unplanned shifts in the bureaucratic body politic. To be sure, you could start the week with a meeting with official X and end up receiving a phone call from his/her successor the next day. Yet, what was clear to me was that the vision laid out by the two leaders in a lengthy joint statement following PM Modi's meeting with President Trump in February, guides the different contours of the relationship at the functional level. Notwithstanding the game of political catch between outlandish tweets and measured official responses, the guidelines for those moving the relationship across government and the private sector are more or less clear. First, there is a concerted effort to realise outcomes in the strategic technology partnership between the two sides. Under the banner of TRUST (Transforming the Relationship Utilising Strategic Technology), the administrative State and technology companies between the two countries are working towards outcomes to do more on pharmaceuticals with the view to de-risk the production of key ingredients from China; fuse infrastructure partnerships between firms invested in the present and the future of Artificial Intelligence (AI); and actively looking for ways to cooperate on extracting and processing critical minerals. The latter needs work, but the zest to find the right compact is real. Second, American private sector actors are preparing the ground to sell different kinds of reactors to meet India's nuclear energy needs. They are, at this time, hoping that the proposed legislative changes to the Indian Civil Nuclear Liability Act 2010 streamline liability clauses in consonance with global standards — delinking liabilities on suppliers and operators. Further, they remain hopeful that changes to India's Atomic Energy Act would allow private sector participation to meet India's nuclear energy needs. This is a top priority for the White House and the US President. This was made clear in several exchanges. This is 'unfinished business' following the conclusion of the 2008 US-India Civil Nuclear Agreement, as one Washington insider put it. My own sense is that progress on this front is almost as important as the conclusion of the first tranche of the trade deal between the two countries. The first tranche of this deal needs to be completed by July 9, when the US President's 90-day pause on 'reciprocal tariffs' ends. Officials suggested that there is a fair chance that the first tranche of the deal with India will be completed by this deadline. 'The trickier parts will come later,' they made plain. Third, efforts across bureaucracies in Washington D.C. that deal with India are almost uniformly focussed on the Quad Leader's Summit in October or November, which provides an opportunity for another bilateral between the two leaders. 'Deliverables' is the name of the game. Yet, at least some of the deliverables need to be real. The ongoing process is less about padding a joint statement and more about searching for right-sized deals. There is a fire in the system to make something happen by the time the leaders meet, including a considerable push to realise new AI infrastructure partnerships. 'India and Brazil are the two most important countries for the US when it comes to data centres,' as one technocrat stated. 'We need to get this right on both sides', the official made plain. In the US, this would mean producing revised rules for export controls that make it easier to access chips from the US into India. In turn, India will possibly need to negotiate certain guarantees to make sure that the chips are not off-shored. Moreover, there is a significant push to deregulate the data centre market in India, and streamline processes to encourage the expansion of AI infrastructure in India. None of this will be easy. Deregulation takes time. Negotiating guarantees can be cumbersome and is a process that cuts across several administrative buildings in and across New Delhi and other Indian states. If Indian officials conclude that data centre investments are an advantage for India, this is the bureaucratic work that will be required to realise this unique moment. It is exactly why structures like TRUST were created, for top political leaders to monitor progress on crucial initiatives. The enthusiasm for investments and partnerships will not last long. This also might be kept in mind. This is a zero-sum play. In sum, while there is little doubt that Munir, Pakistan, Twitter exchanges, and the politics that shape these expressions and incidents to an extent inform the current state of US-India ties, at times exercising officials on both sides, it is also plainly clear that the functional relationship — which produces material results — is one that is working to produce outcomes, and not without the direction of the political leadership. Rudra Chaudhuri is director, Carnegie India. The views expressed are personal.
