Ha ha, Goes with this one nitter.cf/rravi/status/212233719β¦
Damn. Unbelievable feats are normalised as inevitable next steps. No they are not.
This normalisation must end.
30
Damn. Unbelievable feats are normalised as inevitable next steps. No they are not.
This normalisation must end.
If the AI models are as good, then the latest claude or openai would already overcome the CUDA limitation & being stuck with GPU compute & unleash optimal CPU archs or unlock CUDA so efficiently that NVDA is not as relevant.
82
Without getting into any personalities, it's a huge win for Democracy. The system self-corrects.
The anti-war Immune system of Americans have primarily kicked in and they have chosen the Alternative. One can say the anti-crazy immunity too.
78
This is an all round good thread from @myprasanna
BitMela Jaipur Talk: Roaring 20s are back again? Same script 100 years later.
Huge wealth comes from riding a new wave. What the 1920s can teach us about AI's boom, the upcoming 2029 crash, and crypto.
1
80
Trump was too much outward focussed to missing the onward.
Texas flipping is a blow the Reds will not come out of.
52
This is how it's going to be. @karpathy giving the "sutra" the "formula" for the app!
AI can regenerate it.
58
Any of these to trillion $ in months?
Cybersecurity has become more critical with AI.
Top 5 pure-play cybersecurity firms by recent market cap (approx Oct 2026):
1. Palo Alto Networks (PANW): ~$300B mkt cap, ~$11.5B rev, ~$307M net income
2. CrowdStrike (CRWD): ~$257B mkt cap, ~$5.4B rev, ~$45M net income
3. Fortinet (FTNT): ~$100B mkt cap, ~$7.9B rev, ~$2.3B net income
4. Cloudflare (NET): ~$100B mkt cap, ~$2.5B rev, ~-$206M net loss
5. Zscaler (ZS): ~$39B mkt cap, ~$3.4B rev, ~-$63M net loss
70
The contradiction of needing to put millions of people through manual laborious tasks for making money & thus economy versus automating the jobs so machines can do 'em
How can economy work when millions will be rendered jobless?
Intelligence & labour both will be decimated.
46
Tokenmaxxing that propelled the AI revenue is ending with 2026. This was done coz
1) most companies didn't know what AI could do, hence had a free run
2) Big AI cos were filing IPO and needed to boost revenue & pushed for tokens.
Now this model changes for good.
1
47
A question that am trying to figure out.
At 90,000 $ per capita GDP, why should USA still have China or others manufacture products for them. Why can't they have their own production of most products? Do they really need that 15-20% cost savings?
Other than unwillingness to work on factory floors or do some services, the cost savings can't be it.
But given the inequality and disparity between rich and poor, many Americans would be willing at a decent minimum wage (say 25$ an hr). So why not do it all within USA?
Same goes for EU at 47-50k per capita GDP? But EU has external energy dependency and entirely self sufficient as USA could be.
37
Profile of a 17yr old Dario.
Indian kids & kids everywhere can get inspired. He went on to build Anthropic.
Anthropic IPO news will cloud this.
103
Engagement is everything!
T-cell engagers are a new to me but they clearly are getting the job done here.
I captured this video in my lab to show the power of medicines called T-cell engagers.
Left: T cells move in and out of a cancer tumoroid, largely ignoring it.
Right: Add a T-cell engager, and the T cells attack. Once they recognize their targets, they tear through the tumor.
3
88
Algal DHA omega-3 supplements are a bust.
No good for brain (it was expected to improve brain health & no benefits for hearts either. Earlier studies point β« Afib.
The brain study PreventE4 (Yassine et al., eBioMedicine 2026): 365 adults aged 55-80 at dementia risk (47% APOE4 carriers, low fish intake) took 2 g/day DHA or placebo for 2 years. CSF DHA rose ~17% at 6 months (reached the brain, independent of APOE4). No difference vs placebo in memory scores or hippocampal volume at 24 months.
Heart refs: VITAL/STRENGTH (no CV benefit for mixed omega-3s); 2021 meta-analysis (higher AFib risk); REDUCE-IT (positive for pure EPA)
1
85
2027 will be the year of mass layoffs. Progressively worse into 2030s.
Then humanity will realise there is a new norm.
Wars and news around it and about our economies will keep us occupado. But the grounds below our feet would be shifting meanwhile.
AI, robotics & automation
4
161
Plastics are the yet to be banned DDT for humanity.
nature.com/articles/s41586-0β¦
43
The damages done by a single bureaucrat is immense. This has screwed up an entire civilisation for past 175 years.
In 1835, Macaulay dismissed the medical teaching in India's Sanskrit and Arabic colleges as "medical doctrines which would disgrace an English farrier".
33 years later, an official British publication explained why Indian drugs mattered: they could save the colonial government money.
The book was the Pharmacopoeia of India, published in London in 1868 under the authority of the Secretary of State for India.
Its preface wanted medical officers to become "in a great measure independent of costly imported articles", producing "a considerable annual saving to the State".
There is a separate list inside: 40 medicinal articles made official for India that were absent from the British Pharmacopoeia. Among them were neem, catechu, guduchi and isabgol.
The preface also acknowledged that some indigenous drugs had already entered the British Pharmacopoeia itself.
And the people behind this work deserve names.
- Moodeen Sheriff, a Native Surgeon in Madras, compiled medicinal-drug names in twelve Indian languages. The editors described it as "a work of immense labour, reflecting the greatest credit on the intelligence and industry of the compiler". His supplement appeared in 1869.
- The contributors included Bholanauth Bose, Odoy Chund Dutt and Kanhai (Kanny) Lall Dey. Dey's own book on indigenous drugs, published in 1867, was cited in the official volume.
Understand here in the colonial record: contempt in an education-policy argument, followed by official recognition of Indian remedies, Indian expertise and the savings they could deliver.
Sources: T. B. Macaulay, Minute on Education, 2 February 1835; E. J. Waring, ed., Pharmacopoeia of India, 1868, Preface pp. v-ix, contributors pp. xiii-xiv, medicinal-article list pp. xv-xvi; Moodeen Sheriff, Supplement to the Pharmacopoeia of India, 1869.
Photo 1: page xv lists medicinal articles included in the Indian Pharmacopoeia but absent from the British Pharmacopoeia. Highlighted: guduchi, neem and catechu. Pharmacopoeia of India (1868)
Photo 2: the actual 1868 preface, page x. It credits Moodeen Sheriff and his 12 language catalogue, and describes reducing dependence on costly imports. I have highlighted the relevant lines.
3
57












