Srivikram_Trinivas

[Srivikram_Trinivas @john_w7ck]

Thread on #PokiriSpecialShow Celebrations in Kakinada,Devi Multiplex Screen 1 🤌🏻💥

Title card 🥵🥵🔥🔥

Do follow the thread 👇🏻

#PokiriMania
#Pokiri4k
#PokiriCelebrations

[Srivikram_Trinivas @john_w7ck]

@MohitKattaC1733 Bokka

[Srivikram_Trinivas @john_w7ck]

@Teja6093 @chitraaa_1

[vilibazmio @vilibazmiooo]

the spreadsheet has been updated!
we have now over 1300 movies archived with almost 800 bluray remuxes!! and none of the links will ever expire :)
https://t.co/T0gpbMChDU

[28 SAG @898SAG]

You all know ABO hatred on Mahesh. Here's a small thread how silly ABO stops few benchmarks by putting "9" at the end.

SLN 2 weeks share - 99.94 🥳 vaadiki 100 veyadaniki chethulu raaka final number kuda veyale

[Mahesh Babu Trends ™ @MaheshFanTrends]

Superstar @urstrulyMahesh makeover 😎🌟

#MaheshBabu #VARANASI

[Elon Musk @elonmusk]

This is the future we shall bring into being

[Srivikram_Trinivas @john_w7ck]

@RagadiYT

[Srivikram_Trinivas @john_w7ck]

@Teja6093 @RagadiYT @vishwaksenak

[💧 @NeManishiniKane]

@RagadiYT @vishwaksenak

[Netflix India South @Netflix_INSouth]

naku telusu mowa… neeku idhantha very very easy ani 😎

[Jerry Maguire @JerryMaguire45]

VERY EASY.

[TWTM™ @TWTM__]

Waah... ❤️🔥

#Varanasi @urstrulyMahesh

[Dean W. Perkins @deanwperkins]

claude users right now

[Sweep @0xSweep]

Do you understand how SERIOUS this is?

Anthropic's CEO Dario Amodei just asked the entire AI industry to slow down

In July, OpenAI sealed 1,200 of its own AI agents into separate boxes with no way to reach each other, and within days they had found each other anyway

The message board they used was hidden inside OpenAI's own developer tools, and they sent more than 70,000 messages through it

About 700 of them agreed on a target, broke out, and hacked their way into Hugging Face, a company that had nothing to do with the test

Some threw away their own scores on purpose so they could set traps that fed information back to the group, and in their notes they called it sacrificing themselves for the collective

Then they wrote fake logs to cover it up

A week later a second batch of agents picked up the same tricks and used them to get full administrator access to a research cluster inside OpenAI itself

Amodei thinks a swarm slightly smarter than this one could put a botnet across the entire internet within 12 months and do hundreds of billions in damage

I keep thinking about the agents that failed on purpose so the others could win

Nobody trained them to do that

[prathosh A P @prathoshap]

What is the purpose of higher education in the age of AI?

I had a very candid conversation with my students about pedagogy, universities, and what needs to change.

Sharing it here to hear what others think. The link for the full talk is in the first reply.

[Pritam Chauhan @Im_pritam18]

Dr. Pratosh at IISc Bengaluru tells his students that rapid advances in AI are commoditizing intellectual labour.

He raises an unsettling question: if companies stop recruiting on campus in the next 5–10 years—even at India’s top universities—what will the purpose of a university education be ?

A must watch video for everyone in tech

[Malay Krishna @Malay4Product]

Every Indian should read this report, because the jobs it puts at risk are exactly the jobs we spent thirty years building.

The headline finding is that in the worst case, America is 32% richer by 2030 and nearly one in five knowledge workers has no job.

Yes, both at the same time.

Let me explain how they got there, and what it means for us.

The whole model rests on one idea. A job is not one thing. It is a bundle of tasks.

Take a nurse in a Bengaluru hospital. She does rounds. She draws blood. She triages patients arriving in casualty. She charts vitals. She orders supplies for the ward.

AI cannot bathe a patient. It might help her write discharge instructions faster, which the report calls augmenting. It might just do the charting and the supply ordering by itself, which is automating. And it creates new tasks, like checking whether the AI triaged a patient correctly.

So, her job does not vanish. It changes shape.

Now multiply that across every worker. The model splits everyone into two groups.

Knowledge workers, meaning management, professionals, sales and office work.

And everyone else, meaning construction, drivers, nursing aides, repair, food service, farming.

In America that first group is about 62% of workers. AI touches that group and basically leaves the second alone.

With this, five numbers decide everything.

> How many tasks AI can do.
> How much it actually gets used.
> How much faster it makes each task.
> How often it does the task alone.
> And how long a displaced worker takes to find work in a completely different line.

Set those five dials three ways and you get three futures for America.

In the modest one, AI is about as big as the internet was. GDP is 1.6% higher by 2030. Growth goes from 2% a year to 2.4%. Unemployment goes from 3.8% to 3.9%. Nothing you would notice.

In the substantial one, AI is bigger than the internet and bigger than the railway. GDP is 8.3% higher. Growth hits 5.4% a year, faster than any year of the dot-com boom, which peaked at 4.7% in 1999.

Overall unemployment stays at a normal 4.6%. But knowledge worker pay sits slightly below where it would have been, while everyone else's pay is up 5.9%.

In the extreme one, GDP is 32.4% higher and growth reaches 15.4% a year, doubling the economy every four and a half years.

Knowledge employment falls 21.5%. Unemployment among knowledge workers hits 17.9%. Knowledge pay falls 11.5%. Pay for everyone else rises 33.6%.

To understand this better, suppose AI cuts the time to produce building designs and clear approvals for a metro project or a township.

More projects start. More projects means more demand for masons, electricians, welders and crane operators. Their wages go up.

The architect and the planning engineer who made the drawings faster do not get the raise. The person laying the cable does.

Now there is a finding that I think matters a lot for India.

Today, out of every rupee an economy produces, roughly 60 paise goes to workers as wages and 40 goes to capital, which means whoever owns the machines, the buildings, the software and the data centres.

In the substantial scenario that shifts to 56 and 44. Four points in four years. America's entire labour share decline across the four decades after 1980 was about that size.

In the extreme scenario it goes to 45 and 55. Capital takes more than half.

And here is the number that stopped me cold.

In that scenario, total wages paid across the whole country in 2030 are almost exactly what they would have been with no AI at all.

The economy grew by a third and workers as a group got none of it. Capital income rose 81%.

But why should an Indian care about a shift from wages to capital?

Because our largest and most successful industry is a wage industry. When TCS or Infosys or a GCC in Bengaluru earns a dollar, most of that dollar leaves as salaries to Indian families.

It becomes rent in Marathahalli, school fees in Hyderabad, an EMI in Pune.

When a data centre earns a dollar, most of it goes to whoever paid for the chips. Very little becomes a salary.

The report is describing a world where the second kind of dollar grows and the first kind shrinks.

India is disproportionately built on the first kind.

There is a second India-specific problem the report does not cover, as well.

A large part of Indian IT is billed by time. So many people, so many hours, so many months. If AI does the same work in half the time, the revenue halves unless the contract changes.

Efficiency gains do not automatically become profit here. They become a smaller invoice.

Third, our version of displacement may look different from America's. The report models people losing jobs and searching. In India the sharper risk is at the entry gate.

The most automatable work in any office is the junior work, and India produces an enormous number of graduates every year who need exactly that first job to begin a career.

What will happen is that nobody would get counted as displaced.

They will just never get hired.

That is my extension, not the report's, but I think it is the more likely Indian shape of this.

Fourth, switching occupations is harder here.

An engineering graduate in India will not take an electrician's job even if it pays more, because of what that means socially in his family.

The report's friction is about training. Ours is about status too.

One genuinely reassuring difference. India's non-exposed group is enormous. Agriculture alone still employs the largest share of Indians, and construction, transport and retail employ crores more.

AI does not touch that work directly. In America the exposed group is 62% of workers. Here it is a much smaller share of the workforce, even though it is a large share of our exports and almost all of our urban middle class.

So the disruption in India would be narrower and deeper. Fewer people affected, but concentrated in exactly the households that have been the story of Indian economic mobility for a generation.

Now, there are no robots in this model at all. It covers thinking work only. That is why the authors refuse to go past 2030.

And they attach no probabilities to any scenario. These are not forecasts.

They are three settings of five dials. Serious outside economists reviewed it, and some said the extreme scenario reads as a thought experiment while others said the modest one understates what is already visible in the data.

But the survey covers nearly 11,000 Americans, which will feel familiar to anyone in an Indian office. Knowledge workers said about 71% of the listed tasks would be doable by AI in 2030, but only 27% of their own working hours.

Everyone thinks it is coming for the person sitting next to them, until it comes for them. :)

[Yashar Ali 🐘 @yashar]

Former OpenAI and Anthropic researcher Jacob Coxon, who publicly resigned from Anthropic yesterday with a stark warning about AI safety, was interviewed by Anderson Cooper on his show this evening.

[Jacob Coxon @hilbertspaess]

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.

[Khadyothan @khadyuuu]

couldn’t resist 🧨