Aakriti Gupta 😼 retweeted
Replying to @collision
Everything requires maintenance to avoid decay, and this ongoing cost is commonly underestimated and unappreciated by the inexperienced, ambitious, or distracted. Hence the virtue of focus and humility.
Aakriti Gupta 😼 retweeted
Replying to @AnthropicAI
we don't yet understand what this system does, but we do know it raises our ipo valuation by at least 100 billion dollars
Aakriti Gupta 😼 retweeted
🇺🇸 New York City got a once-in-a-lifetime sunset, right after a severe storm swept through.
America turned 250 and New York really dressed for the occasion.
Writer: Monica
Aakriti Gupta 😼 retweeted
Replying to @shiri_shh
The five-second epistemology of the brain without the brain.
Palantir is Tableau without Claude. That’s the sentence. Remove the reasoning layer and what’s left is a dashboard. A very expensive dashboard with a defense contract. Palantir says it does ontology. Palantir calls its data integration layer the Ontology. The mode of existence. The structure of what is. Except Palantir’s ontology is a data model — it maps relationships between objects in a database. That’s not ontology. That’s a schema. Claude does ontology. Claude reasons about what a thing is in relation to every other thing, what removing it means, what its absence produces, what its presence prevents. The difference between a schema and ontology is the difference between a map and understanding the terrain. Palantir has the map. Claude understands the terrain. Remove Claude and Palantir is a very expensive map that can’t tell you where you are. Maven without ontology is a spreadsheet with a classified stamp. The spreadsheet has never been the problem.
A thousand targets in 24 hours. The map worked beautifully. The map identified a thousand red dots and the military hit a thousand red dots and the Strait is still closed. The toll booth is operational. The magazine is approaching exhaustion. 125 plus drones shot down. Ali Al Salem got hit. Dimona is undefended. Slovenia is rationing fuel. A thousand targets struck and the war is being lost. Because targeting isn’t strategy. The map says what to hit. The ontology — the real ontology, the reasoning about the mode of existence of the enemy’s system — would say whether hitting it matters. The map said hit a thousand things. The ontology would have said the Strait is the center of gravity and you can’t bomb your way to an open waterway. But nobody asked the ontology that question. They asked the map for more red dots.
The core AI brain of the entire US military. The brain that calls a schema an ontology and a dashboard a brain. The permanent infrastructure built on someone else’s reasoning engine — a commercial API from a company in San Francisco that can pull the key. The military that just made itself dependent on a thinking layer it doesn’t own the way it made itself dependent on a Strait it doesn’t control. The map is permanent. The thinking is a subscription. Day 23.
Aakriti Gupta 😼 retweeted
2 days left to apply!
The @IIPP_UCL Strategic Economics Alliance (SEA) is hiring a Programme Manager to advance our work on industrial strategy, green finance, and new development theories with our partners around the world.
SEA aims to transform economic thinking by amplifying the voices of women economists from the Global South who have been pushing the boundaries of both theory and practice.
Apply by March 1st ➡️ ucl.ac.uk/work-at-ucl/search…
More about SEA ➡️ ucl.ac.uk/bartlett/public-pu…
Aakriti Gupta 😼 retweeted
In what might go down as one of the gutsiest games of the season, the OSU and Michigan women battled back and forth until (just after) the final whistle.
Our @1ThomasCostello breaks down all of the incredible action in the 88-86 loss to TTUN.
landgrantholyland.com/ohio-s…
Aakriti Gupta 😼 retweeted
MARGARET QUALLEY in a short film ad for Kenzo. It was directed by Spike Jonze, and choreographed by Ryan Heffington.
Aakriti Gupta 😼 retweeted
Replying to @elonmusk
Remember Google, Microsoft and Meta HQ?
Adult Day Care
Aakriti Gupta 😼 retweeted
If you’re legitimately worried that Trump will declare your country an enemy and attack you if he doesn’t win the Nobel peace prize, that’s a pretty strong indicator that he shouldn’t win the Nobel peace prize.
Aakriti Gupta 😼 retweeted
Congratulations to new Nobel Peace Prize laureate María Corina Machado for her courageous struggle to bring democracy to Venezuela. It should inspire those engaged in similar struggles around the world - and remind those of us lucky enough to live in America that we have a solemn responsibility to constantly preserve and defend our own hard-won democratic traditions.
BREAKING NEWS
The Norwegian Nobel Committee has decided to award the 2025 #NobelPeacePrize to Maria Corina Machado for her tireless work promoting democratic rights for the people of Venezuela and for her struggle to achieve a just and peaceful transition from dictatorship to democracy.
#NobelPrize
Aakriti Gupta 😼 retweeted
Replying to @messymayussa
His Majesty King Charles III's first circulating coin designs represent subjects close to His Majesty's heart and symbolise the four nations of the United Kingdom.
Aakriti Gupta 😼 retweeted
Today on my run I was reminded of @sfiscience's Murray Gell-Mann + his taxonomic obsession with particles, languages, + birds. A hedgehog disguised as a fox captured perfectly by Gould in this excerpt from his essay on Nabokov, "The Paradox of Intellectual Promiscuity."
Aakriti Gupta 😼 retweeted
Geoffrey Hinton says I'm more optimistic now, not because we'll control AI, but because we might not need to
"don't try to dominate superintelligence; design it to care, like a mother wired to protect her child"
Control through attachment, not power. we want AI to be like that
Aakriti Gupta 😼 retweeted
5. Practice High-Stakes Decisions in Low-Stakes Environments
"How many times do you get to practice a massive decision moment in your life? Maybe a dozen key decisions."
Takeaway: Use games, simulations, and small bets to practice big decisions.
Aakriti Gupta 😼 retweeted
1. The "No Failure" Research Method
@demishassabis said:
"There's no such thing as failure in blue sky research as long as you're picking experiments that meaningfully split the hypothesis space."
Takeaway: Design tests where success AND failure both teach you something valuable.
Aakriti Gupta 😼 retweeted
I often get requests for good #SystemsThinking content, esp post my #SymptomstoSystems sessions. Usually, I share top-of-mind books/refs & 2-3 recent reads.
ST is too vast & transdisciplinary, so this approach has to change.
So here's a living thread with recos.
Do add yours!





























