Sreeraj

A message for us! #environment #nature #conservation #sustainability #ecology

Earth Day gives us an important opportunity to reflect on the state of our planet – our only home. Just a few weeks ago, NASA’s groundbreaking Artemis II crew shared a new image of Earth, reigniting the spirit of awe and wonder that inspired the first Earth Day. Traveling farther into space than any humans ever have, the crew looked back at our planet and captured an image as it set behind the moon. This image, called Earthset, mirrors the famous Earthrise photograph from the Apollo 8 mission that sparked the modern environmental movement more than 50 years ago. I remember seeing the Earthrise image for the first time and feeling called upon to protect the fragile, pale blue dot we call home. I hope this new image will inspire a new generation of leaders to feel the same way. As Artemis II astronaut Christina Koch said, reflecting on her crew’s mission: "We will inspire. But ultimately, we will always choose Earth. We will always choose each other." Indeed, we must always choose the health and safety of the place we call home and the people we share this home with. On this Earth Day, and every day, I encourage you to commit to creating a cleaner, safer, and more sustainable future for us all.

Los Angeles becomes first major school district to limit classroom screen time bbc.in/4tU7plG

🦔A new class of AI startups is bragging about spending more on AI compute than on human salaries, a trend being called tokenmaxxing. Swan AI's CEO posted on LinkedIn that his four-person team's $113,000 monthly Claude bill was something to be proud of, arguing the token spend replaces their sales, engineering, support, and legal functions. Meta has an internal leaderboard called Claudenomics tracking employee token usage as a productivity metric. The trend is being driven partly by the goal of building companies with as few employees as possible, with investors pushing founders toward autonomous companies with little to no headcount. What's largely unaddressed is whether the AI spend is generating returns, whether token prices will stay at current levels as frontier AI companies remain deeply unprofitable, and what happens when the AI gets it wrong. My Take The circular financing problem I've been covering all year is visible in miniature here. OpenAI and Anthropic are losing money on every token sold because compute costs exceed revenue. Startups are treating those subsidized tokens as a replacement for human capital. The investors funding those startups are often the same investors funding the frontier AI companies selling the subsidized tokens. Everyone is booking the spend as evidence of growth while the underlying economics remain unresolved. A human employee with domain expertise compounds in value over time, builds institutional knowledge, and doesn't get 48% more expensive when GPU spot prices spike. Token costs are going up as the compute crunch tightens. The startups celebrating their AI bills today are building cost structures that depend on prices frontier AI companies cannot sustain without continued investor subsidy. When that subsidy adjusts, the math looks very different, and the founders who bragged about their invoices will be the ones explaining to investors why their unit economics never worked. Hedgie🤗

**TATA CONSULTANCY SERVICES JUST INCINERATED 12,200 INDIANS TO FEED AMERICA'S AI BLOODLUST WHILE THEIR $25 BILLION REVENUE FUNDS THE OFFSHORE EXECUTION OF 400,000 MORE JOBS** 12,200 people. Middle management. Senior architects. Team leads who built the backbone of global IT. Executed because American companies decided AI + 3 junior contractors in Hyderabad costs less than experienced Indian engineers TCS made $25 billion last year. Profit margins at 25%. Stock up 18% But they're "optimizing for AI-augmented delivery models" Translation: we're firing the Indians who were supposed to replace the Americans, so AI can replace both The beautiful fucking irony? These are the same engineers who spent 2 decades automating American jobs Now they're getting automated themselves I'm hearing the entire $283 billion Indian IT sector is planning synchronized cuts. Sources saying 500,000 jobs targeted for "AI transition" by 2027 Infosys, Wipro, HCL all running the same playbook. Fire the seniors. Keep the juniors who cost $12K annually and can run Claude prompts The contractors replacing them? $8K per year. Fresh college grads who think prompt engineering is a career Meanwhile every American CEO is jerking off to "10x productivity gains through offshore AI partnerships" What happens when you offshore the offshore? When you automate the automation? They're about to find out **If you're an engineer anywhere in the world, you're next**

This is a FRAUDULENT paper, AI-generated. My name was used as an author and I had nothing to do with it, never saw it until today e-pubmed.co.uk/journals/digit… The "Editors" Angelo Rossi Mori, David Mensah, and Zarnie Khadjesari should be reported.

@halairporttrfps Kempapura Main Road to Yemalur fully jammed. Please help. @blrcitytraffic @BlrCityPolice @CPBlr @DCPTrEastBCP @bellandurutrfps

Palantir and the New Order: Neoliberalism is dead. Say hello to Techlordism thepoint.com.au/opinions/26042…

Back to books - Sweden's schools give up digital learning bbc.in/41CvrFS

AI chatbots misdiagnose in over 80% of early medical cases, study finds ft.trib.al/cRifiz2

@RemaNagarajan Like BPOs, data centres are 'growth', 'development', investment, modern, ease of doing business...what not?

Imagine not even quantifying the water being used for data centres when cities are increasingly water stressed

Non-disclosure agreements (NDAs) are typically signed between *private* individuals or businesses to prohibit the sharing of sensitive private information. Seeking NDAs in the pursuance of a national tiger survey, as is being done at Wildlife Institute of India, is ridiculous

One month of using only induction stove to do basic cooking for two people, electricity bill jumped by ~₹2200

Bengaluru's green cover has fallen from 78% to 6%, its natural cooling systems are gone, and the city is baking. #KarnatakaNews #Bengaluru deccanherald.com/india/karnatak…

So, hardcore pro-LDF Malayalam news channel has taken a protection cover. bestmediainfo.com/mediainfo/tele… #Kerala #Journalism

🦔A 30-year Oracle veteran posted on LinkedIn suggesting the company's layoffs may have targeted employees with outstanding stock options, noting that many of the best colleagues were cut. Former employees on Blind and TheLayoff.com made similar claims, with some saying they were laid off shortly before vesting dates. Laid-off employees immediately forfeited unvested stock. Oracle declined to comment. A senior manager confirmed the cuts were not performance-based, saying the company lost some of its most experienced, critical people. Oracle simultaneously hired a new CFO with a $26 million stock package on top of a $950,000 base salary. My Take The Oracle story keeps adding detail and none of it is flattering. Timing layoffs before vesting dates to claw back stock compensation that employees earned is not new in tech, but doing it at this scale while simultaneously awarding the incoming CFO $26 million in stock and filing thousands of H-1B petitions makes the calculus unusually visible. The people who built Oracle's products for decades lost unvested compensation they had been working toward. The new financial officer starts with a package that exceeds what most of those employees would have seen in a lifetime. Nobody has confirmed an algorithm targeted vest dates specifically and that should be said clearly. What has been confirmed is that the cuts were not performance based, that they hit some of the most experienced people, and that the timing coincided with upcoming vesting events for many of those affected. Whether that was deliberate targeting or an outcome of cutting by seniority and compensation level, the result is the same. The people who absorbed the cost of Oracle's AI infrastructure bet are not the people who made it. Hedgie🤗

“The most painful feelings and the most stinging emotions are those that are absurd – the longing for impossible things, precisely because they are impossible; nostalgia for what never was; the desire for what could have been; regret over not being someone else” — Pessoa

Will AI slop kill human newsletters? Every day, I encounter new 10-minute-read long newsletters that are clearly LLM-generated word salads. LLM-generated word salads are based on content scraped from the internet. They are, by definition, mediocre. Most people will not notice that it is a human-sounding LLM word salad. It takes only a few minutes to create very long and human-sounding LLM-generated word salads about any topic. People will be inundated by many long, LLM-generated, seemingly sophisticated, but mediocre newsletters. Human authors cannot keep pace with LLM-generated newsletters. Human newsletters die.

"The intellectual is, quite simply, a human being who has a pencil in his or her hand when reading a book." –George Steiner (one of the great literary critics)

“It is often tragic to see how blatantly a man bungles his own life and the lives of others yet remains totally incapable of seeing how much the whole tragedy originates in himself, and how he continually feeds it and keeps it going.” — Carl Jung