Porsches, triple salaries & $100 million offers: AI startups go wild with hiring offers as tech giants scoop top talent
BENGALURU: When Karan Vaidya, co-founder of AI infra startup Composio, promised a Porsche to anyone who referred a product engineer who joined their San Francisco team and stayed for three months, it sounded like a spiel.
"Not kidding," he wrote in the now-viral X post, complete with a photo of a black Porsche model he had in mind.
You Can Also Check:
Bengaluru AQI
|
Weather in Bengaluru
|
Bank Holidays in Bengaluru
|
Public Holidays in Bengaluru
Just hours later, Roy Lee, co-founder of Cluely, another early-stage AI startup, chimed in with his own no-nonsense offer. "I'll triple your base salary," he posted, as he searched for a founding designer. "No questions asked."
The timing is uncanny. This same week, Microsoft reportedly poached 24 AI researchers from Google DeepMind, including Amar Subramanya, the former engineering head of Gemini team who is now a corporate VP at Microsoft AI.
Meta hired three DeepMind experts who had helped build a language model that achieved gold medal-level performance in the International Math Olympiad.
Big money can hurt small AI teams, say experts
Meta has also been aggressively staffing its new Superintelligence unit, reportedly offering compensation packages of over $100 million to lure top researchers from OpenAI and Anthropic.
"There's a real risk that we're overpaying for momentum and mistaking it for durability," said Manav Garg, co-founder and managing partner at Together Fund.
"High-visibility compensation moves can bring in mercenaries rather than missionaries. That's dangerous for early-stage companies where cultural fit and long-term belief in the vision are as critical as technical skill."
For many investors, the current frenzy feels inevitable. "Talent is IP," said Avijeet Alagathi, founder of Shastra VC. "Capital solves for little in this space, so it's being used to buy and retain talent."
Thiyagarajan Maruthavanan, co-founder at Upekkha, described the market as unforgiving. "In artificial intelligence , the fourth player doesn't even matter," he said. "The people who can orchestrate systems at scale are reaching infinite price."
For now, in a winner-takes-all environment, founders are betting that conviction, like a Porsche, can still turn heads.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
2 days ago
- Time of India
Meta will now let candidates use AI in interviews: Is this the future of hiring in Silicon Valley?
In a move that might signal a seismic shift in how talent is evaluated, Meta has confirmed that it is actively testing AI-enabled interviews where coding candidates can use an artificial intelligence assistant during the assessment process. The update was first reported by 404 Media, which accessed an internal memo titled 'AI-Enabled Interviews—Call for Mock Candidates.' The document invites existing employees to volunteer as test subjects for mock interviews involving AI support, noting that this model is 'more representative of the developer environment that our future employees will work in.' It's not just an experimental idea. Meta appears to be laying the groundwork for a new industry norm, one that acknowledges how AI is already deeply embedded in real-world workflows. From help to hiring: Why Meta is betting on AI tools Confirming the internal pilot, a Meta spokesperson told 404 Media, 'We're obviously focused on using AI to help engineers with their day-to-day work, so it should be no surprise that we're testing how to provide these tools to applicants during interviews.' This is the same company where, just months ago, CEO Mark Zuckerberg stated on the Joe Rogan Experience podcast that he expected AI agents to soon function as 'midlevel engineers' capable of writing code for the company. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Pierce Brosnan's Wife Lost 120 Pounds - This Is Her Now Undo 'Over time,' he added, 'we'll get to a point where a lot of the code in our apps is actually going to be built by AI engineers instead of people engineers.' So why test humans in a way that doesn't reflect that reality? That's the core logic behind Meta's pilot. By allowing AI during interviews, the company is signalling that it's less interested in testing raw memory or speed, and more in evaluating how future employees might collaborate with AI systems under real conditions. It's a big pivot from conventional wisdom and it's catching on across the Valley. A glimpse into the future: Cluely, Columbia and ethical grey zones If Meta is at one end of the curve, Chungin 'Roy' Lee is the outlier who saw it coming just a bit too early. In early 2024, Lee, a then 21-year-old computer science student at Columbia University, developed Interview Coder, an AI tool that could discreetly assist candidates during coding interviews. It captured screen activity, picked up audio cues, and suggested real-time responses. For users, it was a lifeline. For institutions, it was a line crossed. Columbia suspended Lee for a year, citing violation of academic integrity. The scandal could have ended the story but it didn't. Lee moved to San Francisco, rebranded the product as Cluely, and turned it into a stealth-mode AI assistant offering undetectable support not just in interviews, but also exams, meetings and sales calls. By mid-2025, Cluely had raised over $20 million in funding, including a $15 million Series A round led by Andreessen Horowitz. Lee remains defiant about his journey. Speaking to the Associated Press, he said: 'Everyone uses AI now. It doesn't make sense to have systems that test people as if they don't.' While Cluely sits in a controversial corner of the AI productivity movement, Meta's recent decision suggests the narrative is shifting. Rather than penalising candidates for using AI tools, some companies are now building interviews around them. The interview cheat code or new literacy? The trend raises an important question: If AI is going to be an integral part of professional life, should interviews continue to pretend it doesn't exist? For many Silicon Valley employers, the answer may be no. Hiring teams are increasingly interested in how candidates interact with AI, rather than how they perform in isolation. As coding environments become more collaborative with GitHub Copilot, Google Gemini Code Assist and Meta's own Code Llama now common in developer stacks, pure unaided performance is losing relevance. Meanwhile, allowing AI during interviews might make it harder to distinguish between strategic assistance and dependency. It could also widen the gap between candidates trained in AI-integrated environments and those coming from more traditional setups. What does this mean for future applicants? For candidates preparing to enter the tech workforce in 2025 and beyond, Meta's pilot could change how they approach not just interviews but learning itself. It normalises the presence of AI in high-stakes environments. It also redefines what companies consider 'cheating.' The emphasis now appears to be shifting from what you know to how you adapt. In that landscape, the ability to prompt, parse and co-create with AI may soon become just as important as any core coding language. As Chungin Lee's story shows, the tools that once got students suspended may very well become the tools that help them get hired. Meta's pilot is not just a test of software; it's a preview of a job market where human-AI collaboration isn't an advantage, it's the baseline. TOI Education is on WhatsApp now. Follow us here. Ready to navigate global policies? Secure your overseas future. Get expert guidance now!
Time of India
5 days ago
- Time of India
AI startup CEO who Amazon got suspended from college starts new ‘employee dating' policy: ‘Please message me directly…'
Image credit: LinkedIn Artificial Intelligence startup Cluely has introduced a new perk for employees: a matchmaking program that rewards staff for successfully referring dates for their colleagues. In a post shared on professional networking site LinkedIn, the company's CEO Chungin "Roy" Lee unveiled this "companywide policy". Under the new policy, employees will receive $500 (around Rs 43,000) for each successful date referral. The incentive is "stackable," meaning an employee can earn multiple bonuses by setting up different colleagues. The 21-year-old CEO was suspended from Columbia University after Amazon pressured the school over his AI tool, which provided undetectable real-time assistance in exams and interviews held in virtual settings. Lee even used the tool to succeed in interviews at major tech firms before publicly showcasing it, leading to revoked offers and a university investigation. What Cluely CEO said about the new employee dating policy Announcing the policy, Lee wrote: 'Cluely now offers $500 cash referrals for dates. Here's the memo I sent out today: "@everyone NEW COMPANY-WIDE POLICY: If any employee refers a date for any other employee (that they are happy with), the referring employee will receive a one-time cash bonus of $500. This is infinitely stackable. ie if Ben refers a date to Neel whom Neel is happy with, and then subsequently refers a date to Brandon, whom Brandon is happy with, Ben would receive $500 + $500 = $1,000" by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like The Simple Morning Habit for a Flatter Belly After 50! Lulutox Undo Please message me directly if you find any members of the Cluely team attractive. Dating is, and will always be, an important part of our culture until we are all happily married.' Lee, who was a computer science student at Columbia University, was suspended last year for violating the university's Academic Integrity Policy. The AI tool he developed, which at that time was called Interview Coder, helped users with coding interviews. Using it, Lee even secured internship offers from tech giants like Amazon, Meta, and TikTok. He later shared a YouTube video demonstrating its use during an Amazon interview, which reportedly led Amazon to complain to Columbia University. In response, Lee said he had already declined the offer and defended his tool as a critique of outdated interview practices. 5 Tips to Get the Best Deals during sale on Amazon, Flipkart and other online websites AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Economic Times
5 days ago
- Economic Times
Gemini's gold-medal AI model is live, announces Sundar Pichai
Agencies Google has rolled out a version of its artificial intelligence (AI) model Deep Think in the Gemini app, its chief executive, Sundar Pichai, said in a post on X. Gemini 2.5 Deep Think achieved the gold-medal standard at the International Math Olympiad (IMO) competition held between July 10-20 on Australia's Sunshine Coast. The results marked the first time that AI systems such as OpenAI and Google's Gemini crossed the gold-medal scoring threshold at the IMO for high school students. However, when it comes to solving complex math problems, the current model takes hours to reason and solve user queries. In a blog post, the company said the currently available model is capable of achieving bronze-level performance on the 2025 IMO benchmark, based on internal the company said, the Gemini 2.5 Deep Think model has been rolled out to a small group of mathematicians and academics to take early feedback and enhance the model's research and inquiry. Parallel thinking techniques Google's Deep Think model utilises parallel thinking techniques, an approach that lets Gemini generate multiple ideas at once, weighing potential solutions to solve complex problems and considering them simultaneously before giving the final idea is to integrate creative thinking capabilities into the AI model by exploring different hypotheses and giving it 'thinking time.'Gemini positions Deep Think as a powerful tool for researchers, as it can reason through 'complex scientific literature' to highly complex mathematical problems and along with coding. State-of-the-art performance Gemini 2.5 Deep Think achieved state-of-the-art performance across benchmarks like LiveCodeBench V6, which measures competitive code performance, and humanity's last exam, which measures expertise in different domains, including science and math. AI taking charge According to a Reuters report published on July 22, Junehyuk Jung, a math professor at Brown University and visiting researcher in Google's DeepMind AI unit, said AI is less than a year away from being used by mathematicians to crack unsolved research problems at the frontier of the field. CEO Sundar Pichai had also mentioned that AI has brought about a significant platform shift, allowing people, businesses, and communities all over the world to access decades of research. Elevate your knowledge and leadership skills at a cost cheaper than your daily tea. US tariff hike to hit Indian exports, may push RBI towards rate cuts Is Bajaj Finance facing its HDFC Bank moment? Tata Motors' INR38k crore Iveco buy: Factors that can make investors nervous Trump tariffs: End of road or a new journey ending Russia reliance? Stock Radar: PI Industries stock showing signs of momentum; takes support above 50-DEMA – time to buy? Long-term investing: Volatility, even threats, have limited shelf life; 5 large-caps from different sectors with upside potential of up to 38% These large- and mid-cap stocks can give more than 21% return in 1 year, according to analysts Weekly Top Picks: These stocks scored 10 on 10 on Stock Reports Plus
