Nothing Linear

Markets. Businesses. Convexity. Nothing is linear | |

[Nothing Linear @NothingLinear01]

@NITIAayog has released a roadmap targeting India's bioeconomy at $691B by 2035 and $2.6T by 2047 (vs $195.3B today)

What's interesting is that the report names national missions, a funding mechanism, and a candid assessment of the bottlenecks
@DrJitendraSingh #BioEconomy

[Nothing Linear @NothingLinear01]

Absolutely! Sooner, we will just stop using software and start delegating our lives to agents

[Vijay Shekhar Sharma @vijayshekhar]

Just as by the end of 2025 we experienced coding and voice as superb AI use cases.
By the end of 2026, we will have super-capable AI assistants—personal agents irresistibly good at performing all the tasks we do on our phones and computers.

Apple will announce OS access to trusted agents, just as it allowed other apps to handle calling/SPAM management.

Android will have to open up and allow third-party agents at the same level as Gemini.

When we enter 2027, many of us will be using agents as few of us were using smartphones in 2010s.

This will take AI usage to the masses. More common people (not-just-nerds) will experience superb productivity gains.

Agents for common business work will be as normal as was, an intranet (for internal ops) or a website/app (for customer connections).

Just as today we are identified by phone number, email, or social handle, we will have our own agents that interact with the agentic world around us.

In 2027, organizations and businesses will rush to launch agents, just as they launched websites and apps in the early internet and mobile-internet days.

The world will see a leap in productivity like never before. We won’t tap touchscreens or click on computer, as much.

Just as tapping took over clicking (mobile apps ate desktop internet), agents will eat apps and our experience of interacting with computing will change for good.

2027 will be massification of AI.
2026 will seal the lead of machines generating codes over programmers.
And, we will start our journey of AGI.

🤖 #ai_bull.

[Nothing Linear @NothingLinear01]

@nvidia 's PAIR is an interesting step toward distributed personal AI compute

It lets multiple PCs, GPUs and Macs on the same network share inference workloads, routing requests to whichever machine has capacity

Could this be how personal AI compute is increasingly built?

[Nothing Linear @NothingLinear01]

#Makr Microsystems just raised ₹10.2 Cr to build #nanoscale 3D #metrology for inspecting structures beneath chip surfaces

And that becomes increasingly important as #chipmaking moves toward advanced packaging, hybrid bonding and chiplets

[Nothing Linear @NothingLinear01]

India's AI buildout is creating demand across an entire infrastructure stack:

GPUs → Data Centres → Networking → Cooling → Power → Cloud → Semiconductor infrastructure

The companies enabling this stack could become an important part of India's AI investment story

[Nothing Linear @NothingLinear01]

#Yotta Data Services is targeting a $1–1.5B IPO in Jan–Mar 2027, in what could become one of the more interesting listings in India’s #AI infrastructure ecosystem

The company plans to use the proceeds to buy #GPUs , repay debt and expand its sovereign cloud infrastructure

[Nothing Linear @NothingLinear01]

#Yotta is exploring structures where partners finance #GPU purchases through #SPVs , with revenues generated by those GPUs being shared before ownership eventually transfers to Yotta

That could help Yotta expand compute capacity while reducing the upfront capital burden.

[Nothing Linear @NothingLinear01]

The #nuclear renaissance is being priced as a reactor story

The chokepoint is fuel — specifically #HALEU , the enriched #uranium >50% of new SMR designs need

#Russia and #China make it at commercial scale. Western capacity is 3+ years from filling the gap

A thread 🧵

[Nothing Linear @NothingLinear01]

We’ve categorised every corporate filing into relevant buckets — from capacity expansions and acquisitions to order wins, management changes and regulatory events!

The idea is simple: turn a filing feed into a way to identify opportunities!

Explore it on https://t.co/I89g5fvRP7

[Nothing Linear @NothingLinear01]

Does the West scale HALEU capacity fast enough for 2028-2030 SMR deployment — or does the first wave run on prepaid Russian fuel via waiver extension?
Either path re-prices the fuel-cycle layer

[Nothing Linear @NothingLinear01]

The #nuclear renaissance is being priced as a reactor story

The chokepoint is fuel — specifically #HALEU , the enriched #uranium >50% of new SMR designs need

#Russia and #China make it at commercial scale. Western capacity is 3+ years from filling the gap

A thread 🧵

[Nothing Linear @NothingLinear01]

The picks-and-shovels stack:

#Mining : Cameco, Kazatomprom, NexGen

#Enrichment : Centrus (LEU) — only Western commercial HALEU pure-play

#Fuel form: Standard Nuclear (STDN) — only listed TRISO pure-play

[Nothing Linear @NothingLinear01]

#Defence exports reached ₹38k cr in '26.

The portfolio is expanding beyond #BrahMos into #drones, #radars, #EW, counter-drone systems and precision weapons

India’s next def exp opportunity may be less about flagship platforms and more about building a broader technology stack

[Nothing Linear @NothingLinear01]

@jeevanpatwa Here's a list of a few #Robotics companies globally for us to track

https://t.co/7QS6aXpAgG

If you know of any good firms in the robotics space (listed or unlisted, Indian or global) that are not on this list please suggest their names, we will add them and track them closely

[Nothing Linear @NothingLinear01]

Hardware iteration speed in Chinese humanoid robotics is fast enough

It will be interesting to watch China's development in the software/behavorial layer too!

[Nothing Linear @NothingLinear01]

Chinese #humanoid robot ran the 100m in 9.39 seconds at the World Humanoid Robot Games in Beijing on Saturday

Usain Bolt's human record: 9.58s (2009).

This is even more interesting given the learning curve effect🧵

[Nothing Linear @NothingLinear01]

China currently produces majority of the world's humanoid robots

Events like this one are running alongside the '26 World Robot Conference, where roughly 3k robotics products were shown — both are functioning as demonstrations of manufacturing and R&D capability