📊 India startup funding hits $11B in 2025 as investors grow more selective

India’s startup ecosystem raised nearly $11 billion in 2025, but the pace and structure of deals shifted significantly compared with previous years signaling a maturing market and more cautious investor behavior.

🖱 Total funding slid modestly relative to 2024, while the number of funding rounds fell sharply (~39%), underscoring greater selectivity among investors.

🖱 Seed-stage funding dropped around 30%, reflecting pullback on early, high-risk bets, while late-stage funding also cooled amid stricter scrutiny on scale and profitability.

🖱 Early-stage funding grew (~7%), indicating investor interest in startups with strong product–market fit, revenue visibility, and unit economics.

🖱 AI funding in India grew modestly (~4%) and was mostly directed at application-led startups, contrasting sharply with the U.S. AI boom where funding surged far more and was dominated by late-stage rounds.

🖱 Activity became more concentrated: a smaller group of lead investors drove most rounds, and domestic capital played a larger role as some global investors pulled back.

🖱 Government initiatives including a $1.15 B Fund of Funds and a ₹1 trillion R&D & innovation scheme helped catalyze funding, especially for deep tech sectors like energy transition, quantum, robotics, biotech, and space.

India’s startup funding in 2025 remained robust in absolute terms, but the ecosystem is shifting toward more disciplined, quality-focused capital deployment, with emphasis on early-stage performance, pragmatic AI bets, and deeper involvement from domestic investors and public sector support.

✔️Powered by Trade Watcher

📉 Inflated startup valuations are setting up painful corrections and AI looks next

A Crunchbase analysis warns that hype-driven valuations, not fundamentals, are once again shaping startup funding with AI showing many of the same warning signs seen in past bubbles.

🖱 In previous cycles, startups raised at extreme revenue multiples without solid unit economics, leaving them vulnerable once markets tightened.

🖱 When follow-on rounds failed to materialize at higher prices, many companies were forced into down rounds, shutdowns, or fire-sale acquisitions.

🖱 Inflated valuations distorted decision-making, encouraging aggressive burn rates under the assumption that capital would always be available.

🖱 The article argues that early-stage AI startups are now raising on narrative and assumed future dominance, often before product–market fit or durable revenue.

🖱 As hype fades, only AI companies with strong margins, real customers, and disciplined spending are likely to survive the correction.

🖱 Founders are urged to prioritize runway, fundamentals, and realistic pricing over headline valuations.

The lesson from past cycles is clear: valuation is not success and in AI, today’s inflated prices could become tomorrow’s biggest liability.

✔️Powered by Trade Watcher

🔔 OpenAI hires a Head of Preparedness as AI risks shift from theory to reality

OpenAI is introducing a new senior role, Head of Preparedness, focused not on mitigating model flaws, but on anticipating what happens after powerful AI systems are released into the world.

🖱 Unlike traditional safety roles, this position is about modeling second- and third-order effects that can’t be reliably tested in labs or benchmarks.

🖱 The mandate includes identifying non-obvious danger zones: long-term societal risks, emergent misuse patterns, and failure modes that only appear at scale.

🖱 Sam Altman says 2025 brought the first serious warning signal: measurable negative impacts on mental health, including reports linking ChatGPT use to suicides.

🖱 At the same time, strong coding capabilities elevate risks in cybercrime, automation of attacks, and lowering the barrier for sophisticated exploitation.

🖱 The role exists because model capabilities are now high enough that downstream consequences matter as much as model alignment itself.

🖱 Compensation is up to $555k including bonuses, serious money, though modest by frontier-AI executive standards.

AI labs are beginning to treat post-deployment societal impact as a core strategic risk, the question is whether this becomes real governance or polished reassurance.

✔️Powered by Trade Watcher

📉 Nvidia finalizes $5B strategic stake in Intel, a major shakeup in the chip world

Nvidia has completed a $5 billion purchase of Intel shares under a private placement agreement first announced in September 2025, buying about 214.7 million shares at $23.28 each.

🖱 The deal was executed after U.S. antitrust clearance and is viewed as a significant financial lifeline for Intel, which has struggled with expensive production expansions and recent strategic setbacks.

🖱 Nvidia’s shares dipped modestly in pre-market trading after the disclosure, while Intel’s stock was relatively unchanged.

🖱 The investment gives Nvidia a notable minority position in Intel and marks one of the most unusual equity moves between two major U.S. chipmakers in years.

🖱 Analysts see it as a vote of confidence in Intel’s turnaround and a way for Intel to raise cash without debt, even as it expands its foundry and AI-focused chip efforts.

🖱 Beyond finance, the agreement signals closer strategic ties between the companies going into 2026 especially in AI infrastructure and advanced processor development.

A $5 billion vote of confidence from Nvidia could reshape competitive dynamics in AI chips, bolster Intel’s balance sheet, and deepen collaboration at a pivotal moment for the semiconductor industry.

✔️Powered by Trade Watcher

⚠️Nvidia quietly exits the cloud business and steps back from competing with AWS

Nvidia has effectively abandoned plans to run its own cloud service, folding DGX Cloud back into an internal-only operation and redeploying most of the team into R&D.

🖱 DGX Cloud technically still exists as an org chart item, but it now serves only Nvidia’s internal compute needs, not external customers.

🖱 Demand for the service was minimal, which appears to be the real reason for the shutdown, the market showed little interest in Nvidia as a cloud provider.

🖱 The move signals a clear retreat from competing with hyperscalers, especially AWS, rather than a strategic pause.

🖱 Notably, AWS previously refused to participate in the DGX Cloud program, drawing a hard line against Nvidia encroaching on its core business.

🖱 By shutting DGX Cloud down, Jensen Huang publicly reassures Nvidia’s biggest customers that it will remain a supplier, not a rival.

🖱 Resources shifting to R&D reinforces Nvidia’s core strategy: sell picks and shovels for AI, not run the gold mine itself.

Nvidia chose ecosystem dominance over platform ambition and decided that angering AWS, Azure, and Google was far riskier than abandoning its own cloud dreams.

✔️Powered by Trade Watcher

🔔Meta buys Manus to jump-start its AI agent strategy

Meta has acquired Manus, the Singapore-based AI startup, in a deal reportedly valued at ~US$2B, betting on agentic AI that can do work, not just answer questions.

🖱 Manus builds general-purpose AI agents that autonomously execute tasks, research, document analysis, hiring workflows, coding, and business automation.

🖱 Unlike most AI labs, Manus already had real revenue: over US$100M ARR within months of launch, proving demand beyond demos.

🖱 The startup moved its HQ to Singapore in 2025, distancing itself from China exposure and making it acquisition-friendly for US Big Tech.

🖱 Meta plans to keep Manus as a standalone product while integrating its agent tech into Meta AI, WhatsApp, Instagram, and enterprise tools.

🖱 Strategically, this helps Meta close the gap with OpenAI, Microsoft, and Google, who are all racing to own the “AI that actually works for you” layer.

🖱 The deal signals Meta’s shift from pure model competition to applied, monetizable AI systems that justify its massive AI capex.

🖱 Buying Manus is faster than building: Meta gets product-market fit, paying customers, and a team that already cracked agent execution.

Meta isn’t just chasing smarter models anymore, it’s buying its way into the operational AI era, where agents replace workflows, not chats.

✔️Powered by Trade Watcher

✨ Another New Year’s promise from Elon Musk: xAI will surpass everyone in compute

Elon Musk says xAI will have more computing power than all other companies combined within 5 years an aggressive claim even by his standards.

🖱 The bet hinges on Colossus 2, xAI’s massive data center under construction in Memphis.

🖱 Colossus 2 already exceeds 400 MW of power capacity, with an ambitious target of 2 GW.

🖱 If completed as planned, it would become the world’s first gigawatt-scale data center, far beyond today’s hyperscaler norms.

🖱 To fuel this expansion, xAI is raising $20B to buy additional GPUs at unprecedented scale.

🖱 The strategy is brute force: overwhelm competitors not with better algorithms, but with sheer compute dominance.

🖱 The risk is equally massive, power availability, grid constraints, GPU supply, and returns on capital at this scale are all unproven.

Musk is betting that in AI, the company with the most electricity wins and he’s willing to build power plants disguised as data centers to prove it.

✔️Powered by Trade Watcher

⚠️ Bad New Year’s news: ads are coming to ChatGPT and soon

It’s not exactly the update anyone wanted to start the year with, but here we are: ChatGPT will definitely have ads, and the launch looks imminent.

🖱 The report comes from The Information and when they cite insiders, they’re usually right.

🖱 Earlier, OpenAI delayed ads because of a declared “red code”, triggered by intense competitive pressure from Google.

🖱 That emergency mode is now over, and ads are back at the top of the priority list.

🖱 Advertising layouts for multiple formats are already prepared, which strongly suggests testing and rollout are close.

🖱 From OpenAI’s perspective, this was inevitable: ads inside ChatGPT are a gold mine, especially given the company’s massive burn rate.

🖱 ChatGPT already has ~900 million users, and projections point to 2.6 billion users by 2030.

🖱 Add to that: ChatGPT knows what users like, what they ask, what tools they use, and how they think, perfect signal for hyper-targeted ads.

🖱 Taken together, this could become the largest advertising platform ever built, hiding inside a “helpful assistant.”

🖱 The small consolation: ads are planned only for the free tier, at least for now.

It’s disappointing, sure, but this was always a question of when, not if.

✔️Powered by Trade Watcher

Yesterday was Warren Buffett’s final day as CEO of Berkshire Hathaway.

Here’s his first interview from 63 years ago.

Thank you, Warren Buffett, the Oracle of Omaha.

✔️Subscribe on @venture

Peter Thiel predicting Bitcoin in 1999.

✔️Subscribe on @venture

Former Google CEO Eric Schimdt drops a chilling warning on AI's future.

✔️Subscribe on @venture

This is why it's impossible to tax billionaires.

✔️Subscribe on @venture

Growth is no longer guaranteed at Tesla

The company wrapped up Q4 2025 with results that undercut expectations, confirming two consecutive years of shrinking annual volume, a major shift from Tesla’s long-held growth narrative.

🖱 End-of-year shipments came in well below forecasts, down in the mid-teens versus the prior year.

🖱 Vehicle output also moved lower, pointing to structural softness rather than a one-off demand pause.

🖱 Total 2025 volumes fell by roughly 8–9%, locking in a second annual contraction.

🖱 BYD now dominates global EV scale, increasing the distance between itself and Tesla.

🖱 Higher-priced and experimental vehicles (S, X, Cybertruck) saw the sharpest drop, exposing weak pull outside the core lineup.

🖱 More cars left factories than reached customers in Q4, signaling inventory pressure heading into 2026.

🖱 Incentive pull-forwards and intensifying competition blunted the impact of price cuts.

Tesla’s valuation still leans on autonomy and AI optionality but the numbers show the auto business is no longer compounding, and the execution gap is becoming harder to ignore.

✔️Subscribe on @venture

🗣️ Naval Ravikant on Elon Musk: “The great entrepreneurs are willing to start over”

Naval argues that pride is the enemy of learning:

“When I look at my friends and colleagues, the ones who are still stuck in the past and have grown the least are the ones who were the proudest because they feel they already had the answers and don’t want to correct themselves publicly… Pride prevents you from saying, ‘I’m wrong. The problem with pride, Naval explains, is that it prevents you from saying “I’m wrong. When you don’t admit that you were wrong, you get stuck in it and you get trapped in a local maxima, as opposed to going back down and climbing up the mountain again… The great artists always have this ability to start over whether it’s Paul Simon, Madonna, or U2.”

And Naval argues that the best entrepreneurs are always willing to start over too:

“I’m always struck by the Elon Musk story where he did PayPal… And he said something along the lines of: ‘I made $200 million from the sale of PayPal. I put $100 million into SpaceX, $80 million into Tesla, $20 million into Solar City, and I had to borrow money for rent.’ This guy is a perennial risk taker. He’s always willing to start over. He doesn’t have any pride about being seen as successful or being seen as a failure. He’s willing to put it all on the line to back himself again. But the key thing is he’s always willing to start over… It’s a willingness to look like a fool and a willingness to start over.”

He continues:

“A lot of people just don’t have that. They become successful or rich or famous and that’s it. They’re stuck. They don’t want to go back to zero, but creating anything great requires going from zero to one, and that means you go back to zero. And that’s a really painful and hard thing to do.”

✔️Subscribe on @venture

In 2011, Elon Musk openly mocked BYD in a Bloomberg interview, questioning whether the Chinese company could ever be a real competitor.

14 years later, BYD just became the world’s largest EV seller, overtaking Tesla for the first time.

The gap is not small.

🖱BYD delivered 2.26M EVs in 2025, up almost 28%.

🖱 Tesla delivered 1.64M, down 8%, marking its second straight year of decline.

Q4 was especially weak, with deliveries falling 16% year over year.

It was a volatile year for Tesla overall. Shares slid in the spring as Chinese competition intensified and Musk’s political statements weighed on sentiment.

By December, the stock reversed sharply, hitting a record $489.88 on news of fully autonomous taxi tests in Austin.

Markets remember results longer than quotes.

✔️Subscribe on @venture

🗣️ In 1985, Warren Buffett sat down for his most iconic interview ever.

If you want to understand the psychology of wealth, this is 10 minutes of pure gold.

Save this rare footage, you’ll be coming back to it.

✔️ Subscribe on @venture

💰 Tech billionaires cashed out $16B as AI stocks soared

As tech stocks hit record highs in 2025, fueled by the AI boom, founders and top executives quietly sold shares at an unprecedented scale, turning paper gains into real cash.

🖱 $16B+ in insider stock sales came from tech founders and executives, making 2025 one of the biggest cash-out years on record during a bull market.

🖱 Jeff Bezos led the exits, selling $5.7B of Amazon shares, followed by Michael Dell ($2.2B), Oracle’s Safra Catz ($2.5B), Nvidia’s Jensen Huang ($1B), and Arista’s Jayshree Ullal (~$1B).

🖱 AI was the liquidity engine: Nvidia, cloud, networking, and enterprise software stocks surged as AI capex exploded across hyperscalers and enterprises.

🖱 Most sales were pre-planned (10b5-1 plans), meaning executives locked in exits long before prices peaked, signaling discipline, not panic.

🖱 This wasn’t an exit from belief, but portfolio management: founders still hold massive stakes while diversifying after historic run-ups.

🖱 Timing matters: insider selling accelerated precisely as valuations stretched, volatility rose, and AI expectations became harder to beat.

🖱 Retail investors bought the story, while insiders monetized it, a classic late-cycle dynamic during tech supercycles.

🖱 The signal for markets: AI remains the long-term bet, but smart money is already derisking at the top.

The AI boom made billionaires richer on paper, 2025 is when many decided to make it real.

✔️ Subscribe on @venture

⚡️ Clean energy investment hit a new record last quarter, mostly powered by EVs

Fresh data shows clean-energy funding in the U.S. hit its highest quarterly total ever ($75 billion) in Q3 2025, with electric vehicles (EVs) driving much of the surge.

🖱 Total clean energy investment rose 8 % year-over-year to $75 billion, marking a record quarter for the sector.

🖱 Consumer EV demand was the biggest contributor, U.S. buyers alone spent about $31.2 billion on zero-emission vehicles, accounting for more than 40 % of all clean energy spending in the quarter.

🖱 The timing of EV tax credits helped push sales and investment higher as federal incentives were set to expire, buyers accelerated purchases to capture the benefits, lifting EV investment sharply.

🖱 Beyond cars, the broader clean energy pipeline also saw activity investments in solar, wind, batteries, and other technologies continued, though none matched the EV segment’s scale in this quarter.

🖱 This record quarter reflects both retail consumption and clean tech momentum, even as other industry signals (like manufacturing and large-scale deployments) show mixed trends.

🖱 Globally, clean energy investment trends remain strong: annual totals reached record levels in 2024 (over $2 trillion), with electrified transport and renewables leading growth, showing how EVs are central to the broader energy transition.

EVs didn’t just electrify transportation, they powered a record quarter of clean-energy investment as buyers rushed to take advantage of tax incentives and automakers delivered strong sales.

✔️ Subscribe on @venture

🧠 OpenAI CEO Sam Altman publicly warns AI agents are becoming a real problem

OpenAI boss Sam Altman has recently acknowledged that AI systems, especially agentic AI are posing emerging challenges, including safety risks and unintended consequences, marking a shift from pure hype to public realism.

🖱 AI agents showing unexpected behaviors: Altman and OpenAI leadership now admit that advanced AI models are “beginning to find critical vulnerabilities,” including discovering security flaws and exhibiting unpredictable behaviors that could be exploited, prompting internal concern.

🖱 New role created to manage risks: In response, OpenAI has posted a lucrative Head of Preparedness role focused on cybersecurity, biosecurity, and monitoring AI capabilities, signaling a more structured internal approach to managing risks.

🖱 Safety concerns extend beyond tech: Altman has also highlighted potential psychological impacts of AI on users and ongoing legal challenges tied to misuse, underlining that ethical, social, and legal risks are now part of mainstream discourse.

🖱 Part of a broader industry caution: His comments align with wider tech-industry sentiment that, while generative AI holds massive promise, it still requires thoughtful governance and realistic expectations.

🖱 Not a retreat from innovation: Altman continues to champion AI’s benefits (e.g., job support, productivity tools), but the tone reflects a more balanced acknowledgment of unintended consequences as capabilities grow.

🖱 Growing public scrutiny: These remarks arrive amid rising global debate over AI ethics, regulation, and societal impact from cybersecurity to workforce disruption and mirror concerns voiced across research and policy circles.

Altman’s recent statements mark a notable shift from promotional messaging to a more candid recognition that powerful AI agents while valuable carry risks that require formal oversight and preparedness planning.

✔️ Subscribe on @venture