‘What just happened?': IIT Kharagpur student tests ChatGPT o3 on JEE Advanced mock test, taken aback by results
From humanoid robots to self-driving cars to offering relationship advice, artificial intelligence (AI) has become an integral part of several industries, and Sam Altman's ChatGPT has been making waves for quite some time. With multiple new versions, the platform is being continuously refined, impacting professionals across various fields.
Recently, an IIT Kharagpur student tested ChatGPT o3 during her JEE Advanced 2025 mock test, and the results were shocking. In a blog post on software platform Heltar, Anushka Aashvi revealed that the model scored an astonishing 327 out of 360, a result that would have secured All India Rank 4 in the real exam.
Titled 'ChatGPT o3 Scores AIR 4 in JEE Advanced 2025. What Just Happened?', Aashvi shared that she went to great lengths to create a credible exam situation. The model was directed to 'act like a JEE aspirant,' solving each question separately with no internet access and no memory from previous answers. Every question was solved in a fresh chat session to prevent any form of carryover learning.
'We tested the ChatGPT o3 model (which was released on 16th April 2025) on the JEE Advanced 2025 question paper which was conducted on 18th May to ensure that the questions have as much newness for the AI model as possible,' Aashvi wrote.
Despite these constraints, ChatGPT o3 impressed at nearly every step. The platform helped her achieve perfect scores in Chemistry and Mathematics during the second half of the paper, and she lost only a few marks in Physics. The model showed a clear, step-by-step reasoning process, approaching multi-concept questions, advanced calculus problems, and even skeletal chemical diagrams.
'It easily solved lengthy algebra and calculus problems. The model performed remarkably well at combining concepts from multiple chapters to reach a correct solution. It was even able to interpret compounds correctly from their skeletal formulae and solve them correctly,' the student wrote in the blog.
However, ChatGPT o3 did struggle with certain visual and instrument-based questions. Aashvi shared that it failed to accurately interpret a Vernier scale and took nearly 10 minutes to answer a graphical question, only to get it wrong. 'It was not able to understand the Vernier Scale readings. It kept reiterating to get to the solution but took very long and even then gave the wrong answer,' she wrote.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Mint
an hour ago
- Mint
Eye for AI: Stocks recover as tech giants spend big on AI infrastructure — what lies ahead?
Shares of companies that provide infrastructure for artificial intelligence (AI) development bounced back from a tumble earlier this year, as spending by Big Tech restores investor confidence in the tumultuous sector, reported Bloomberg. Goldman Sachs Group tracked two baskets — one tracking AI data centres and electrical equipment stocks, which rose 52 per cent, and another that follows shares of companies that supply power for data centres, which also rose 39 per cent. Both went up from their April lows. Individual standouts in these baskets include Vertiv Holdings Co, which has notched a 94 per cent gain since April 4, as well as Constellation Energy Corp, which went up 75 per cent in that same period. The world's largest technology companies, including Amazon, Alphabet, Microsoft and Meta are continuing to spend big on artificial intelligence, mitigating doubts over whether money will continue flowing to the firms that are essential to AI infrastructure, said Bloomberg. Forecasts for capital expenditures to support AI demand are up by 16 per cent since the beginning of the year, according to Bloomberg Intelligence's Robert Schiffman. 'Earnings season reminded investors that generative-AI doesn't run on buzzwords — it runs on concrete, copper and gigawatts,' CEO of Roundhill Financial Dave Mazza told Bloomberg. Shares of AI infrastructure companies are witnessing an uptick over the business potential of artificial intelligence which sparked a spending spree on data centres to fund the development of AI programs like OpenAI's ChatGPT and Anthropic's Claude, reported Bloomberg. A strong start in 2025 unravelled as worries over competition from China's DeepSeek startup and broader uncertainty over global trade led investors to be wary of the billions of dollars of investment made during the period. Concerns that tech giants such as Microsoft were walking away from data centre projects further pushed the selloff. Investor sentiment improved when President Donald Trump announced the pause of most of the tariffs he had rolled out in early April, which fuelled a rally that pushed the S&P 500 Index near an all-time high hit in February, as per Bloomberg data. Another reason is the latest earnings season which helped investors gain confidence in the sector, as big tech companies indicated they were continuing to lay money out on AI development. Among these was Meta, which signalled that the hundreds of billions in AI spending it had flagged earlier in the year was still on track, said Bloomberg. Recent corporate deals also suggest that spending on AI infrastructure remains intact. Amazon plans to invest $10 billion in North Carolina to expand its data centre infrastructure to support AI and cloud computing technologies, reported Bloomberg. Deputy chief investment officer at Franklin Templeton Investment Solutions, Max Gokhman told Bloomberg that investor confidence in AI could take another hit if the trade war once again heats up and sparks concerns that a downturn in global economic growth will see companies cut on AI spending. 'If the economy falls into a recession, margins will be under pressure, companies will be forced to lay off workers and cut spending on AI,' Gokhman told Bloomberg.
Time of India
2 hours ago
- Time of India
Krutrim announces agentic AI assistant 'Kruti', to launch on June 12
Bhavish Aggarwal 's artificial intelligence (AI) company Krutrim on Monday announced its agentic assistant ' Kruti ', months after the founder first teased the chatbot . In a post on X, Krutrim said Kruti, India's first agentic AI assistant , "listens, adapts and acts proactively, purposefully and in your language". — Krutrim (@Krutrim) Play Video Pause Skip Backward Skip Forward Unmute Current Time 0:00 / Duration 0:00 Loaded : 0% 0:00 Stream Type LIVE Seek to live, currently behind live LIVE Remaining Time - 0:00 1x Playback Rate Chapters Chapters Descriptions descriptions off , selected Captions captions settings , opens captions settings dialog captions off , selected Audio Track Picture-in-Picture Fullscreen This is a modal window. Beginning of dialog window. Escape will cancel and close the window. Text Color White Black Red Green Blue Yellow Magenta Cyan Opacity Opaque Semi-Transparent Text Background Color Black White Red Green Blue Yellow Magenta Cyan Opacity Opaque Semi-Transparent Transparent Caption Area Background Color Black White Red Green Blue Yellow Magenta Cyan Opacity Transparent Semi-Transparent Opaque Font Size 50% 75% 100% 125% 150% 175% 200% 300% 400% Text Edge Style None Raised Depressed Uniform Drop shadow Font Family Proportional Sans-Serif Monospace Sans-Serif Proportional Serif Monospace Serif Casual Script Small Caps Reset restore all settings to the default values Done Close Modal Dialog End of dialog window. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like War Thunder - Register now for free and play against over 75 Million real Players War Thunder Play Now Undo "This is a leap beyond chatbots. More updates on 12th June. Stay tuned!" In April, Aggarwal had posted a shoutout for the team developing the chatbot, saying that the team had worked a lot to make a product as good as the global apps, but better for the Indian users. Live Events He had attached videos and images of the chatbot. Like OpenAI's ChatGPT, the assistant seemed to have three modes: 'Instant' for quick answers, 'Insightful' for reasoning, and 'In-depth' for deep research. 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 This is a key announcement for the company, as ET had reported earlier this month that its other products are not seeing much traction. Founders and developers ET had spoken to saw Krutrim's large language models (LLMs) and cloud offerings as subpar and lacking technical maturity compared to hyperscalers. The Bengaluru-based startup became India's first AI startup to reach unicorn status after it secured $50 million in funding in 2024 at a $1 billion valuation.
Time of India
3 hours ago
- Time of India
Apple WWDC 2025: iPhone maker woos software developers amid AI, regulatory challenges
Apple is facing an unprecedented set of technical and regulatory challenges as some of its key executives are set to take the stage on Monday at the company's annual software developer conference. On the technical side, many of the long-awaited artificial-intelligence features Apple promised at the same conference a year ago have been delayed until next year, even as its rivals such as Alphabet's Google and Microsoft woo developers with a bevvy of new AI features. Those unfulfilled promises included key improvements to Siri, its digital assistant. On the regulatory front, courts in the US and Europe are poised to pull down the lucrative walls around Apple's App Store as even some of the company's former supporters question whether its fees are justified. Those challenges are coming to a head at the same time US President Donald Trump has threatened 25% tariffs on Apple's best-selling iPhone. Apple's shares are down more than 40% since the start of the year, a sharper decline than Google and also lagging the AI-driven gains in Microsoft shares. Apple has launched some of the AI features it promised last year, including a set of writing tools and image-generation tools, but it still relies on partners such as ChatGPT creator OpenAI for some of those capabilities. Bloomberg has reported that Apple may open up in-house AI models to developers this year. Live Events But analysts do not believe Apple yet has what technologists call a "multi-modal" model—that is, one capable of understanding imagery, audio and language at the same time—that could power a pair of smart glasses, a category that has become a runaway hit for Meta Platforms. Google said last month it would jump back in to this category, with partners. 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 Such glasses, which are far lighter and cheaper than Apple's Vision Pro headset, could become useful because they would understand what the user is looking at and could help answer questions about it. While Apple has focused on its $3,500 Vision Pro headset, Google and Meta have seized on the smart glasses as a cheaper way to deploy their AI software prowess against Apple in its stronghold of hardware. Meta Ray-Bans all sell for less than $400. Analysts say Apple needs to answer that challenge but that it is not likely to do so this week. "I'm not trying to replace my phone—this is a complementary thing that gives me more world context, because it's got a camera and it sees what I see, and I can talk to it in natural language," said Ben Bajarin, CEO of technology consultancy Creative Strategies. "Apple is not positioned to do that." To be sure, Apple's rivals are not decisively ahead in smart glasses. Anshel Sag, principal analyst with Moor Insights & Strategy, said Meta's Ray-Bans still lack some features and Google has not yet landed its "Gemini" model in a mass-market pair of glasses yet. "Meta has the undisputed lead, but Google is catching up fast and probably has the best-suited AI for the job," Sag said. "Vision Pro is great, but it's a showroom product that developers can use." But Bob O'Donnell, CEO of TECHnalysis Research, said it remains far from clear that smart glasses will gain wide acceptance. O'Donnell also said it is not certain that Apple is at any particular disadvantage if it partners with a company such as Google, OpenAI or even a smaller firm like Perplexity for core AI technology. So far, O'Donnell said, there is not yet strong evidence that consumers are basing major hardware-purchasing decisions on AI features. "There's an argument to be made that it's OK that (Apple) is behind because, except for the bleeding edge, most people don't care," O'Donnell said.
