Hari

[RoninTraders @RoninInvestors]

Everyone's talking about AI datacentres, Semiconductors, Optics, Memory, but what actually are they, well...

[Ricardo @Ric_RTP]

Google just quit the AI race on purpose, and it is about to make MORE money than everyone still running it.

4 of the most cited AI researchers alive walked out of Google in a single afternoon.

Jeff Dean, the man who built the systems Google runs on, gone after 27 years. Sanjay Ghemawat, his longtime partner, gone. Oriol Vinyals, a Gemini co-lead, gone. Quoc Le, a Google Brain co-founder, gone.

That same day, Demis Hassabis stepped back from running DeepMind. Hassabis co-founded the lab, won a Nobel Prize for AlphaFold, and had been the face of Google AI for a decade.

The stock dropped 5% within hours. Analysts called it a brain drain. Headlines called it the day Google fell behind.

But turns out that's completely wrong, because the numbers underneath tell a completely different story:

Google is not trying to win the frontier model race anymore. It looked at where the money is and walked toward it.

Gemini, Google's flagship model business, generated about $12 billion in annual revenue last quarter. That is the entire payoff from competing head to head with OpenAI and Anthropic.

Now look at the other number.

By the end of 2027, Google Cloud is projected to do over $73 billion selling AI infrastructure to other companies, plus another $120 billion selling its TPU chips. That is roughly $200 billion of external sales at high margins, against a $12 billion model business.

Google understood that the frontier race is the expensive part while selling the shovels is the profitable part.

And the customers buying those shovels include Google's own rivals.

Over 20% of Google's TPU shipments for 2026 and 2027 are going to Anthropic, one of the two labs supposedly beating Gemini. Google now makes money every time Anthropic trains a model designed to crush Google's OWN product.

Cede the frontier, own the layer underneath it, and collect a toll from everyone racing across the top.

The researchers leaving is the symptom of a company that already decided models are not where it wins.

Jeff Dean said it himself on the way out. He told the New York Times that leaving a public company gives him room to make decisions "not necessarily in the company's purist financial interests."

Read that from Google's side:

The people who wanted to chase the science left, because Google is now optimizing for the FINANCIAL interest.

Gemini 3.5 Pro is running months behind, with staff blaming low morale. DeepMind's comms, legal, and marketing teams are being folded into Google proper. A former manager told the Guardian the era of DeepMind as an independent lab is over.

None of that reads as failure once you see the strategy.

Yet Wall Street is pricing this as Google losing.

The parallel that should worry the frontier labs:

If open weight models keep compressing the price of inference, being the best model stops being a business. It becomes like semiconductor fabrication, strategically vital and financially brutal, a race you win and still lose money running.

Google is the first giant to admit that.

The company that invented the transformer just handed the frontier to OpenAI and Anthropic, and positioned itself to get paid on every model both of them ship.

Those labs will be burning billions to stay one benchmark ahead, and Google will be cashing in hundreds of billions from it.

The model business is actually just a race where everyone loses.

Apple understood that from the get-go and never joined the race, Google understood it now and left it to OpenAI and Anthropic.

Who will go bankrupt first?

[Naval @naval]

New podcast on vibe coding - A Return to Code.

A Return to Coding 00:20

The Personal App Store 03:17

Vibe Coding Is a Video Game with Real-World Rewards 06:22

Pure Software Is Uninvestable 10:33

A Place for Each Model 14:22

AI Is Eager to Please 17:57

Why Math and Coding? 22:10

The Beginning of the End of Apple’s Dominance 24:17

Coding Agents As Customer Service Reps 27:55

[Ravi Prakash Official @raviprakash_rtv]

I’m a Telugu journalist based out of Hyderabad and I’m writing this on behalf of EVERY SOUTH INDIAN STATE.
India needs to pay URGENT attention !!

This is not delimitation. This is political punishment.
Punishment for controlling population.
Punishment for investing in education.
Punishment for doing exactly what the nation asked us to do.

For decades, South Indian states acted responsibly , built human capital, strengthened public systems, and contributed disproportionately to India’s economy.

And now, the reward?
Fewer voices in Parliament.
Let that sink in.
States that governed better are being pushed to the margins, while political power shifts purely on population numbers.

This is not federalism. This is a structural imbalance in the making.
Is India still a Union of States or are we heading toward permanent dominance by a few, at the cost of others?

Representation cannot come at the cost of balance.

Democracy cannot mean silencing those who performed better.

If delimitation happens without safeguards, it won’t just redraw constituencies , it will redraw the idea of India itself.
SOUTH INDIA IS NOT ASKING FOR PRIVELEGE.
It is asking for fairness.

@narendramodi @PMOIndia @ncbn @AndhraPradeshCM @revanth_anumula @TelanganaCMO @siddaramaiah @CMofKarnataka @mkstalin @CMOTamilnadu @pinarayivijayan @CMOKerala

[Nav Toor @heynavtoor]

🚨 Claude Code costs $200/month. GitHub Copilot costs $19/month. Jack Dorsey's company built a free alternative. 35,000 GitHub stars.

It's called Goose.

An open source AI agent built by Block that goes beyond code suggestions. It installs, executes, edits, and tests. With any LLM you choose.

Not autocomplete. Not suggestions. A full autonomous agent that takes actions on your computer.

No vendor lock-in. No monthly subscription. Bring your own model.

Here's what Goose does:

→ Works with ANY LLM. Claude, GPT, Gemini, Llama, DeepSeek, Ollama. Your choice.
→ Reads and understands your entire codebase
→ Writes, edits, and refactors code across multiple files
→ Runs shell commands and installs dependencies
→ Executes and debugs your code automatically
→ Extensible through MCP. Connect it to any external tool.
→ Desktop app, CLI, and web interface. Pick your workflow.
→ Written in Rust. Fast. Lightweight. No bloat.

Here's the wildest part:

Block is a $40 billion company. They built Cash App, Square, and TIDAL. They use Goose internally. Then they open sourced the entire thing.

This isn't a side project from a random developer. This is production-grade tooling from a company that processes billions in payments. Built for their own engineers. Given to everyone.

Claude Code: $200/month. Locked to Claude.
GitHub Copilot: $19/month. Locked to GitHub.
Cursor: $20/month. Locked to their editor.

Goose: Free. Any LLM. Any editor. Any workflow. Forever.

35.3K GitHub stars. 3.3K forks. 4,078 commits. Built by Block.

100% Open Source. Apache 2.0 License.

[jimmah @jamesdouma]

No, this is not competition for FSD anymore than LEGO releasing a Space Shuttle kit is competition for the Falcon 9.

Nvidia has released multiple generations of ADAS development kits and tools for developing ADAS systems. These are not ADAS systems, they are tools to help get started developing an ADAS system. Nvidia has also produced multiple generations of hardware kits that can help a developer get started building the compute framework for an ADAS system using Nvidia silicon. An ADAS demo can be put together pretty quickly using these kits, but a production system cannot - the kit gets you 0.01% of the way to concept for a production system and it doesn't include most of the difficult to understand parts - it just shows what is possible. This latest kit apparently includes the a VLA as the core software architectural component. Using a VLA provides a lot of development advantages but VLAs are compute intensive and not, in their simple form, suitable for a production system.

It would be a good thing for the world if companies picked up these tools and started making a serious attempt to develop ADAS systems and I hope they do. If they were wildly successful they might start fielding them in 5 years and that could help Tesla to displace the billion plus human driven vehicles ten years from now. We need lots and lots and lots of autonomous capable vehicles and Tesla can't build all of them in any reasonable period of time.

There is no scenario in which a company building on top of this new development kit will even slightly dent Tesla's Robotaxi market opportunity. I wish it were that easy - building an FSD like system is still a technically challenging, resource intensive, and commercially fraught task. It's kind of a miracle that any company did it once. It's the thing I'm most grateful to Tesla for.

[Ricardo @Ric_RTP]

Elon Musk just confirmed the most INSANE IPO in history.

SpaceX is going public in 2026.

$1.5 TRILLION valuation. Raising $30+ billion.

That's the biggest IPO ever made. Beating Saudi Aramco's $29 billion record from 2019.

But here's what everyone's missing:

This isn't about space tourism or Mars missions.

Elon is literally about to win the entire AI race.

And 99% of people have no idea how...

Here's the problem killing every AI company right now:

POWER.

Oracle just reported earnings.

They burned through $12 BILLION in one quarter building data centers.

Their free cash flow? NEGATIVE $10 billion.

Revenue missed estimates. Stock crashed 11%.

Microsoft, Amazon, Google all scrambling to find enough electricity for AI training.

The brutal math:

The US generates 490 gigawatts of total power.

AI is projected to need 123 gigawatts by 2035.

That's a QUARTER of the entire electrical grid. Just for artificial intelligence.

Goldman Sachs says AI energy demand could jump 165% by 2030.

There is literally not enough power on Earth to run AI at the scale these companies are promising.

Every data center needs massive cooling systems. Billions of gallons of water per year. Insane energy costs.

And the infrastructure can't keep up.

Elon's solution?

Stop building on Earth entirely.

SpaceX is building data centers in SPACE.

Not a concept. Not 10 years out. Literally starting in 2026.

They're upgrading Starlink V3 satellites to carry AI computing chips.

Each satellite gets 24/7 solar power. No clouds. No night. No weather disruptions. No grid bottlenecks.

And the insane part is that Starship can deliver 300 to 500 gigawatts of solar-powered AI satellites into orbit every single year.

At 300 gigawatts per year, the AI computing power in space would exceed the entire U.S. economy's total electricity consumption within two years.

Just from satellites. Processing in orbit.

While Oracle is begging banks for loans to finish data centers and OpenAI is stuck in circular funding arrangements with Microsoft, Elon already owns everything:

The rockets. The satellites. The launch infrastructure. The AI company (xAI).

He doesn't need to ask utilities for permission.

Doesn't need grid approvals from local governments.

Doesn't need to build nuclear plants or wait for clean energy.

He just launches.

And everyone else is scrambling to catch up:

Jeff Bezos sees it. Blue Origin announced they're building their own orbital data centers.

Google just launched "Project Suncatcher" with plans to deploy AI satellites by 2027.

Eric Schmidt, the former CEO of Google, literally BOUGHT an entire rocket company (Relativity Space) just to compete in this space.

But they're all 3+ years behind Elon.

SpaceX already has 6,000+ Starlink satellites in orbit. The infrastructure is built.

The $30 billion from the IPO?

Going straight into scaling orbital compute.

SpaceX revenue is jumping from $15 billion in 2025 to $24 billion in 2026.

Most of that from Starlink. Now add space-based AI infrastructure on top.

Here's why this matters:

Whoever controls orbital computing controls the AI revolution.

And there's only ONE company on Earth with fully reusable rockets that can launch at the scale required.

Jensen Huang, Nvidia's CEO, called space data centers "a dream."

Translation: Nvidia is screwed if Elon actually pulls this off.

Because if SpaceX succeeds, every AI company on the planet becomes Elon's customer.

OpenAI needs compute? Running on SpaceX satellites.

Google needs more capacity? Renting orbital infrastructure.

Microsoft needs power? Paying SpaceX for launch and compute access.

Elon won't just be in the AI race.

He'll own the entire track everyone else is running on.

The $1.5 trillion valuation sounds crazy until you realize what he's actually building.

It's not a rocket company. It's the infrastructure layer for the next 50 years of computing.

People calling it overvalued have no idea what's coming.

[Peter Girnus 🦅 @gothburz]

Last quarter I rolled out Microsoft Copilot to 4,000 employees.

$30 per seat per month.

$1.4 million annually.

I called it "digital transformation."

The board loved that phrase.

They approved it in eleven minutes.

No one asked what it would actually do.

Including me.

I told everyone it would "10x productivity."

That's not a real number.

But it sounds like one.

HR asked how we'd measure the 10x.

I said we'd "leverage analytics dashboards."

They stopped asking.

Three months later I checked the usage reports.

47 people had opened it.

12 had used it more than once.

One of them was me.

I used it to summarize an email I could have read in 30 seconds.

It took 45 seconds.

Plus the time it took to fix the hallucinations.

But I called it a "pilot success."

Success means the pilot didn't visibly fail.

The CFO asked about ROI.

I showed him a graph.

The graph went up and to the right.

It measured "AI enablement."

I made that metric up.

He nodded approvingly.

We're "AI-enabled" now.

I don't know what that means.

But it's in our investor deck.

A senior developer asked why we didn't use Claude or ChatGPT.

I said we needed "enterprise-grade security."

He asked what that meant.

I said "compliance."

He asked which compliance.

I said "all of them."

He looked skeptical.

I scheduled him for a "career development conversation."

He stopped asking questions.

Microsoft sent a case study team.

They wanted to feature us as a success story.

I told them we "saved 40,000 hours."

I calculated that number by multiplying employees by a number I made up.

They didn't verify it.

They never do.

Now we're on Microsoft's website.

"Global enterprise achieves 40,000 hours of productivity gains with Copilot."

The CEO shared it on LinkedIn.

He got 3,000 likes.

He's never used Copilot.

None of the executives have.

We have an exemption.

"Strategic focus requires minimal digital distraction."

I wrote that policy.

The licenses renew next month.

I'm requesting an expansion.

5,000 more seats.

We haven't used the first 4,000.

But this time we'll "drive adoption."

Adoption means mandatory training.

Training means a 45-minute webinar no one watches.

But completion will be tracked.

Completion is a metric.

Metrics go in dashboards.

Dashboards go in board presentations.

Board presentations get me promoted.

I'll be SVP by Q3.

I still don't know what Copilot does.

But I know what it's for.

It's for showing we're "investing in AI."

Investment means spending.

Spending means commitment.

Commitment means we're serious about the future.

The future is whatever I say it is.

As long as the graph goes up and to the right.

[Sai Deepak J @jsaideepak]

1. Trust the Pakistanis to make a pigsty even out of the @OxfordUnion. And as always, they are genetically incapable of being truthful. So here's the complete story of how this so-called debate played out.

[Alex Veremeyenko @alex_verem]

This might be the most disturbing AI paper of 2025 ☠️

Scientists just proved that large language models can literally rot their own brains the same way humans get brain rot from scrolling junk content online.

They fed models months of viral Twitter data short, high-engagement posts and watched their cognition collapse:

- Reasoning fell by 23%
- Long-context memory dropped 30%
- Personality tests showed spikes in narcissism & psychopathy

And get this even after retraining on clean, high-quality data, the damage didn’t fully heal.

The representational “rot” persisted.

It’s not just bad data → bad output.
It’s bad data → permanent cognitive drift.

The AI equivalent of doomscrolling is real. And it’s already happening.

Full study: llm-brain-rot. github. io

[Abbas Khan @KhanAbbas201]

Been seeing a lot of complaints lately, so I wanted to share more about EF's new structure, which teams have been built out, and who to reach out to for different types of support requests.

First of all, to quote @snapcrackle’ pinned tweet, EcoDev is composed of teams that help users onboard, support builders, steward local communities, and fund public goods so Ethereum stays resilient, human-centered, and useful to the world.

https://t.co/cEdjjldsTg

There are four sub-clusters within EcoDev, but I’ll only highlight Ecosystem Acceleration here since they handle most of the requests the broader community usually asks about:

1. Developer Growth led by @austingriffith with @escottalexander and @binji_x focuses on engaging and supporting next-gen builders.

2. App Relations & Research led by @jchaskin22 with @isha_sangani focuses on accelerating meaningful, user-facing applications.

3. Founder Success led by @adrianmcli with myself and @oel4all focuses on helping founders with fundraising, partnerships, accelerator access, and hosting hackathons to onboard new developers.

4. Enterprise Acceleration led by @davwals with @ashmorgan, @ismiMatthew, and @motypes supports businesses and institutions adopting Ethereum.

So if you're a builder needing dev support, working on a consumer app, an early or mid-stage founder looking for guidance, or an institution wanting to connect with EF, these are the teams to reach out to. They’re all very responsive and supportive.

Lastly, it’s worth remembering that Andre’s criticism is before this new structure. The approach today is much more coordinated, community driven and only a few months old.

[StockMarket.News @_Investinq]

🚨 For 15 years, the Federal Reserve was the most powerful force in global markets.

Now in 2025, that power is slipping away

The financial system is entering a new era.

( a thread)

[Andrew Panella @Longevity_EDU]

The silent killer poisoning you from within:

Insulin resistance.

It's the secret cause of fat gain, heart disease and cancer.

10 hacks to cure insulin resistance:

1. Apple Cider Vinegar

[VatsRohit @KesariDhwaj]

Sargodha, Kirana Hills, Nuclear Weapons and Missile Strike

- Why the US intervened?

- There is ONLY one possible scenario running in my head.

- Pakistanis being Pakistanis, they made a dog and pony show of moving their nuclear assets.

- It could be movement of missile TELs or nuclear weapons form known storage sites.

- Now we know we hit Kirana Hills next to Sargodha.

- Kirana Hills is the motherload of weapons, ammunition and of course, nuclear arsenal.

- Just look at the area -
- It houses a Central Ordnance Depot, Central Ammunition Depot and Pakistan's nuclear stockpile within those hills.

- Which should not be surprising if you consider the following facts -

(1) Kushab Nuclear Plant is ~50km from Kirana Hills as the crow flies. It is part of Pakistan nuclear weapons making infrastructure.

(2) Sargodha Air Base - Home to Pakistan's aerial nuclear delivery platform ->F-16s! Hardly 10km from Kirana Hills in a straight line.

See the map below - The tip of the triangle rests on Kushab Nuclear Plant.

- So, my guess on the sequence of events is something like this -

(1) Pakistanis moved nuclear weapons at a time or in a manner where they knew it would be picked-up by the Americans.

(2) It is very much likely that India also picked-up these events in parallel.

(3) And fired a missile right into their nuclear weapons mothership - No, not to destroy the weapons or make them explode but to tell the Pakistanis that 'We See You' and we're ready to call your nuclear bluff.

(4) But the Americans panicked and well, rest is history.

PS: To see details of where exactly strike hit and details of the area, please see excellent work by @JaidevJamwal on his TL.

[Riley Brown @rileybrown]

Struggle to keep up with all the most powerful AI tools?

Well here is a comprehensive overview:

Sections:
1: AI Tools
2: AI Automations & AI Agents
3: VibeCoding

The Full Mindmap down below 🧵 ⬇️

TIMESTAMPS

A INTRO
00:00 Introduction

B CHAT TOOLS
01:50 ChatTools
01:50 Getting Started with Chat AI Tools
02:27 Exploring Chat GPT's Capabilities
04:11 Creating Custom ChatGPT Projects
07:54 Creating Files (PDF) ChatGPT Projects
11:44 Upload image and Search the web on ChatGPT
15:14 Looking at Google Gemini AI Chat Tool
16:36 Perplexity is another great tool
18:10 This AI Chat Model Can Analyze Videos (Most underrated AI use cases @OfficialLoganK you guys need to hype this up more)
20:16 Recapping Chat Tools

C CREATIVE TOOLS
21:05 Introduction to AI Image Models
22:19 Using Chat GPT-4o for Image Editing
22:30 Using gpt4o images for Business: Repaint Houses
26:36 Edit Parts of an image gpt4o
27:55 Midjourney
28:23 Midjourney is better for exploration
31:25 Midjourney Editor is fun to use

D: CREATING VIDEOS
36:29 Image to Video with Kling and Runway
38:52 Gen4 Turbo
41:22 Kling is Superior
42:00 gpt4o Ghibli Image + Video + Sound
44:58 Ghibli time
49:15 ElevenLabs Sound Effect for pressing button
50:00 Generating Voice for Voice Over
51:04 Getting Music
52:07 Video Editing These together
53:47 Finished the Video - Final Results
55:02 Video Tool Recap
56:27 AI Avatars with Heygen
57:42 Example of good avatar content ROWAN CHUNG
58:43 Pause and Reflect before Automations
59:14 Recap of section 1

E: VIBE FLOWS (AUTOMATIONS AND AGENTS)
59:34 (add here - explanation of workflow automation)
59:34 Zapier Automation
01:04:24 Adding Step in Automation
01:09:45 Stacks, Automations, NOW AGENTS
01:10:29 Deep Research Agent
01:13:11 What is Manus? Let's ask Claude to Diagram it
01:16:25 Prompting Manus
01:18:20 The Future of Agents will look like Online Poker in 2010
01:19:15 WHAT DO WE DO WHEN AGENTS ARE WORKING?
01:19:56 Greg's Theory of "Vibe Marketing"
01:20:44 Manus and Deep Research are done Agenting, let's see what they did
01:23:32 Ok Lets Reflect and Move to VIBE CODING

F: VIBE CODING
01:25:35 VIBE CODING!
01:25:48 Landing Page with Sound, Images, and Video on v0
01:29:07 Generating Video for v0 site
01:31:31 Adding Video and AUDIO to v0 site (ElevenLabs)
01:33:36 Using API's in your Vibe Code Apps (Power Ups)
01:35:26 What makes the best App Idea
01:37:41 Build a simple app with an API with Cursor
01:42:57 Diagraming how the API key works in the Bill Splitter
01:43:25 Structured vs Unstructured text to text API's
01:44:41 Building a mobile app using built in API's from your phone
01:47:31 Creating native share features on ios app
01:49:21 What we talked about in this video (Where to Find the MindMap)

Tools mentioned:
1:ChatGPT, Google Gemini, Anthropic Claude, Perplexity, Grok, GPT‑4o Images / DALL·E 3, MidJourney, Cling 2.0, Krea AI, Runway Gen‑4, Google Veo, Luma Labs, Pika Labs, Heygen, Suno, ElevenLabs
2: Zapier,n8n, Manus, Deep Research, Canva, CapCut, Premiere Pro
3: Cursor, V0, OpenAI API, Replicate, Vercel, Replit, Firebase, Supabase, Claude 3, DeepSeek, Grok, Llama

[Nayini Anurag Reddy @NAR_Handle]

‘400 acres of Hyderabad’s Wilderness Is being Destroyed – This needs National Attention’ #MustRead 🚨

400 acres of thriving wilderness in #Hyderabad, home to 734 plant species, 220 bird species, deer, pythons & centuries old trees, is on the verge of destruction.

Everything you need to know about what’s happening in #Telangana 👇

(1/N)

#SaveKanchaGachibowli
#SaveHCU #SaveHCUBioDiversity #SaveHyderabadBioDiversity

[DogeDesigner @cb_doge]

🚨 Prime Minister Modi on Donald Trump:

Donald Trump has courage. We have a strong bond. During his recent assassination attempt, even after being shot, he remained unwaveringly dedicated to America. His life was for his nation. His reflection showed his America first spirit.

[Hayden Adams 🦄 @haydenzadams]

Just two weeks after shipping v4, we're thrilled to launch @unichain

At Uniswap, we aim to build the best possible experience for onchain markets

Unichain is the next big step ‒ an L2 designed for DeFi. Launching as a stage 1 rollup with 1 second blocktimes, it only gets better from here.

In the months and year to come, we will be launching many new improvements to accelerate blockchain scaling

Flashblocks - a block builder in collab with Flashbots, that will reduce blocktimes from 1s to 0.25s, allow most MEV to be returned to users, and provide additional transparency + decentralization for transaction ordering/sequencing

Seemless interop - we aim for unichain to be the spearhead for advancing what is possible with interoperable L2s, through our partnership with @Optimism as well as our work on intents

Unichain validation network - finally, we plan to continue pushing on L2 decentralization through UVN, which will add an additional layer of economic security and validation on top of the sequencer through staking

In other words, relentless shipping will continue until Ethereum scales and DeFi is bigger than tradfi+cefi combined

In just four months of testnet the network processed ~100M transactions. Now it's live with 80+ projects already building on top (plus @Uniswap + v2, v3, and v4 deployments already live)

I can't wait to see what the community builds on top
🦄⛓️🦄⛓️🦄⛓️🦄⛓️🦄⛓️🦄⛓️🦄⛓️🦄⛓️

[Matthew @MatthewMadera]

TLDR;
Value creation vs Value extraction

On Sonic,
User executes tx
90% of the gas used goes to the dapp (for them to do with as they please)
10% of the gas used goes to validators

On Sonic, our goal is to make the applications $ and keep it within the ecosystem.

On L2's,
User executes tx
tiny % of the gas used goes to L1 to cover settlement
tiny % of the gas used goes to DA solution
large % of the gas used goes to L2 treasury

On L2’s, their goal is to make themselves $ and extract value from their own ecosystem and dump on the Ethereum ecosystem by selling $ETH for USD.

[Free To Speak @FreeToSpeak71]

You bring up a valid point. Taking a cue from Trump we should put "default" into the conversation to shock people into serious considerations of spending. This would get all the holders of Treasuries busy coming up with alternatives that don't wipe them out. And yes granny may hold some bind funds but much bigger fish are going to holding billions and will act to preserve it for them and granny.

Banks, funds, corporations get your politicians to live within a balanced budget or we threaten to default and collapse your house of cards.