Atharva Prabhune

I like cricket, math, investing, reading books, cheesecake and @fcbarcelona

[Arvin @sozirv]

To be called thief, racist, Zionist across millions of posts and news for an entire month, and to reply without a single word only with this pure magic is the greatest story sports will ever tell. With context Messi’s 2026 World Cup is absolutely the best ever ever ever ever

[Fiaz 🍉 @fiaz1899]

There’s this Valdano quote on Messi, how Messi was Maradona of 1986 every day of his career. Not only was he asked to be 1986 Maradona every single week… in 2026 they were still demanding he also be the Mbappé/Haaland of 2026 in the World Cup when he turned 39.

And this freak delivered on both.

The Greatest of All Time. 🐐

[levent @__alpoge__]

hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final

((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)

[Encara @Ankara_M10]

"When Argentina lost" "When Argentina won"

[Jeremy Giffon @jeremygiffon]

https://t.co/srseWsMfr6

[Diego Recalde @DiegoJRecalde]

Messi, el tipo más ganador del planeta, dijo que en la vida se pierde más veces de las que se gana.
Saber eso es la fórmula de la felicidad.

[Kpaxs @Kpaxs]

When you lack a strong opinion of yourself, you become overly dependent on validation.

You need others to confirm what you cannot confirm internally. Compliments become emotional oxygen. Approval becomes addictive. Disapproval becomes devastating. You may end up performing constantly, shaping your personality around what earns acceptance rather than what reflects truth.

And doing so, you abandon your distinctiveness. To avoid disapproval, you smooth out the edges that make you unique. You become more acceptable, perhaps, but less real. Many people become socially palatable at the cost of becoming inwardly disconnected from themselves.

You end up living defensively rather than purposefully.

[Kpaxs @Kpaxs]

If your peace depends entirely on the world behaving gently toward you, then your peace will always be fragile.

But if your peace depends on your ability to process, reject, transform, or outgrow negativity, then you possess a source of strength that circumstances cannot fully control.

It is important to clarify that protecting the mind is not the same as becoming emotionally numb.

A strong person is not someone who feels nothing. Strength is not the absence of hurt. It is the refusal to let hurt become corruption.

You can feel pain without becoming bitter.

You can feel anger without becoming cruel.

You can feel disappointment without becoming hopeless.

You can feel fear without surrendering your direction.

[Atharva Prabhune @atharvakpr]

Nico Gonzalez starting holds the key today. He needs to stretch the pitch, attack Pedro Porro using Messi's L->R switches and help out defensively against Lamine.

Nico also provides a strong aerial threat for Argentina, where Spain is not as comfortable

[Justin Skycak @justinskycak]

People think discipline is punishment. It's actually self-compassion extended through time. You're taking care of a future version of yourself you've never met.

[Atharva Prabhune @atharvakpr]

Witnessing him play has been the privilege of a lifetime

[Justin Skycak @justinskycak]

One of life's great pleasures is becoming the person your younger self would have looked up to.

[Atharva Prabhune @atharvakpr]

If businesses have to deploy AI on prem to protect business secrets, won't IT service companies be the obvious beneficiaries?

[Podcast Alpha @PodcastAlphaX]

@CryptoPunk898 @chamath The biggest reason to move to the cloud was resiliency - because it made sense for the business.

The same way it makes sense for the business to bring AI capabilities on-prem to protect the business.

[Matt Turck @mattturck]

Inside Nemotron and NVIDIA's AI lab: my conversation with Bryan Catanzaro (@ctnzr).

@nvidia is a chip company. So why does it put hundreds of researchers on building AI models - and then give them away for free? We go deep into the Nemotron models, what it takes to build a top AI lab, and the future of frontier AI.

01:33 - Is open source AI catching the frontier?

05:29 - Do closed labs blocking distillation slow open source down?

07:42 - Is the US falling behind China?

10:30 - Why companies actually choose open models

12:39 - A "crazy" 2008 bet: machine learning on GPUs

15:33 - Working with Andrew Ng and Dario Amodei at Baidu

17:41 - Coming back to NVIDIA: DLSS and the birth of Megatron

21:55 - The real reason NVIDIA builds its own models

24:28 - Is Moore's Law really dead?

33:37 - The Nemotron family: Nano, Super, Ultra

35:09 - Built for agents: why NVIDIA bets on speed

36:02 - How you train a 550B model in 4 bits

39:25 - Hybrid Mamba-Transformer, explained simply

42:31 - Mixture of experts, and why NVIDIA built NVL72 around it

47:26 - Why a 1-million-token context window matters

49:26 - Multi-token prediction: how the model predicts 5 tokens at once

52:47 - Multi-teacher distillation: teaching one model from many

58:01 - Where reinforcement learning goes next

01:00:16 - Inside NVIDIA's research org: "the mission is the boss"

01:04:03 - How NVIDIA decides who gets the GPUs

01:10:53 - Why NVIDIA still feels entrepreneurial after 33 years

01:12:58 - Why Bryan doesn't believe in the singularity

01:17:50 - The AI backlash

01:19:18 - The controversial case: open AI is safer than closed

[Atharva Prabhune @atharvakpr]

@ContrarianCurse I'd say:

Theoretical max spend on any ad slot (assuming 100% conversion) = selling price - all variable costs per unit

[Atharva Prabhune @atharvakpr]

@ContrarianCurse The avg. income earner might not spend 10x more on new products.

But if Meta can perfectly target the spend (say I am planning to buy sneakers anyways and Meta recognizes this), then the sneaker companies might be willing to pay much more for a highly relevant ad slot.

[Fireside Alpha @firesidealpha]

https://t.co/QDUZwmvgk0

[Ricky Ho @rickyho_1989]

Hey everyone, I have received permission from my employer to publish this research.

It is a lengthy investment thesis on what I believe will be the next chapter of AI. After months of research, we are turning bullish on the hyperscalers and explain why we believe the market is underestimating where the economics of AI are ultimately heading.

The rest is covered in the X article below. I hope you enjoy reading it, and I look forward to hearing your thoughts and challenges.

[Anjney Midha @AnjneyMidha]

PSA: several folks have asked where they can find the full stanford @CS153Systems '26 lectures

they are uploaded each week on the official @Stanford online youtube channel and we've also created a playlist of them here for easy discovery:

https://t.co/T6zFZ2O7wc

[Atharva Prabhune @atharvakpr]

If model scaling is truly not exponential, we are in a bubble where frontier labs will continuously be squeezed by open source models

[NonsparseOncologist @5_utr]

AI scaling is not exponential. It never was. I fit a logistic model to general capability vs compute across every major model generation — AlexNet to GPT-4. R² = 0.98. The asymptote is not infinity. Never was or will be. Thread 🧵