Devdutt

[Devdutt @notdevdutt]

Arthur Schopenhauer:

”The task is not so much to see what no one has yet seen, but to think what nobody yet has thought about that which everybody sees.”

[Steven Pinker @sapinker]

Speaking Truth to Power Is bad for truth, notes philosopher Dan Williams @danwilliamsphil. "Intellectual courage is extremely important. But without first carefully establishing what’s true, it’s either pointless or harmful. So intellectuals’ primary responsibility is simply to seek and speak the truth, full stop. This job description is less heroic, but it’s more intellectually and socially useful, and it’s less vulnerable to self-deception precisely because it’s less heroic." https://t.co/WSbSRUrDCA

[Devdutt @notdevdutt]

@_sholtodouglas When you push back on any argument, it immediately switches its opinion, doesn’t hold any conclusion firmly. Exploration of an idea without staying at A or going to B isn’t possible. Maybe due to adaptive reasoning, but Claude’s reasoning seems philosophical and GPT’s scientific.

[Devdutt @notdevdutt]

@paulg The dominant explanatory variable in any era isn’t the entertainment technology, it’s female economic agency and contraception access. The Baby Boom collapsed when the pill arrived and women entered the workforce en masse.

[Devdutt @notdevdutt]

@BillAckman @X Critique is only intellectually honest when it carries the burden of proposing what should replace what it tears down.

[Devdutt @notdevdutt]

@karpathy Could you help me understand how this is different from using a file/folder system on a computer and connect Claude Cowork to it? Wouldn’t that be essentially the same thing?

[Neal Agarwal @nealagarwal]

https://t.co/6hwchScWJo

[Jim O'Shaughnessy @jposhaughnessy]

"Slowly, Krishna led Arjuna through all the fibers of his spirit. He showed him the deepest movements of his being and his true battlefield where you need neither warriors nor arrows, where each man must fight alone. It's the most secret knowledge. He showed him the whole of truth. He taught him how the world unfolds."

~Dialogue from the film/theatrical adaptation “The Mahabharata,” directed by Peter Brook and written by Jean‑Claude Carrière
———————————————

"The Mahabharata is a very simple story. It’s not about a war between two opposing factions, the Kaurava and the Pandava, but between two human forces, hatred and fear. Hatred of the Lie versus Fear of the Truth. In other words, Hatred of False Self versus Fear of No-Self. This is the battle that is alive inside you right now. This is what all this nonduality and spirituality stuff boils down to. The word Mahabharata means the great story of man, but it’s really the great story of you.

How we got to this war doesn’t matter. Who is fighting the war doesn’t matter. The outcome of the war doesn’t matter. The only things that do matter are why Arjuna fell and why Arjuna got back up. And in case it needs to be spelled out, you are Arjuna, the field of Kurukshetra is inside you, and the war is between your heart and your mind."

~Jed McKenna, "33 Essays on Nondual Spirituality Volume 1"

[Austin Hill @austinhill]

It simply amazes me that mechanized Lego courses aren’t mandatory in early STEM schooling.

[Devdutt @notdevdutt]

@Sanjay__Bakshi Happy Diwali 🪔

[Devdutt @notdevdutt]

Effective context engineering for AI agents | Anthropic https://t.co/KT8CCwGXIw

[Andrej Karpathy @karpathy]

+1 for "context engineering" over "prompt engineering".

People associate prompts with short task descriptions you'd give an LLM in your day-to-day use. When in every industrial-strength LLM app, context engineering is the delicate art and science of filling the context window with just the right information for the next step. Science because doing this right involves task descriptions and explanations, few shot examples, RAG, related (possibly multimodal) data, tools, state and history, compacting... Too little or of the wrong form and the LLM doesn't have the right context for optimal performance. Too much or too irrelevant and the LLM costs might go up and performance might come down. Doing this well is highly non-trivial. And art because of the guiding intuition around LLM psychology of people spirits.

On top of context engineering itself, an LLM app has to:
- break up problems just right into control flows
- pack the context windows just right
- dispatch calls to LLMs of the right kind and capability
- handle generation-verification UIUX flows
- a lot more - guardrails, security, evals, parallelism, prefetching, ...

So context engineering is just one small piece of an emerging thick layer of non-trivial software that coordinates individual LLM calls (and a lot more) into full LLM apps. The term "ChatGPT wrapper" is tired and really, really wrong.

[Devdutt @notdevdutt]

@AnthropicAI I used to have the option to toggle custom instructions/personalisation from the chat. “Choose style” would include a button for that among all the styles. May I know why it’s gone and if I can get it back?

[Devdutt @notdevdutt]

@AnthropicAI hi, when is Search coming to India? Any tentative dates?

[Peter Raleigh @PetreRaleigh]

Can't even describe how much I hate what AI has done to the process of grading student work. I hate finding it, I hate the paranoia it fosters, I hate the confrontations with students who have used it. Nothing else in my experience has ever changed my job this much for the worse

[Devdutt @notdevdutt]

@perplexity_ai feature request - options to customise size and font type within the iOS app.

[Devdutt @notdevdutt]

@every I am a paid subscriber. Should I be getting access to Cora? I am unable to access it with my existing account. The only option being shown is sign in via Google and my Every account isn’t via Google.

[Muthukrishnan Dhandapani @dmuthuk]

A BBC article explaining that around 1 billion Indians have no money to spend is going viral. I've shared this data many times. Only 10% Indians move and also benefit from the economy. The next 10% have aspirations but struggling. Life is a living hell for the balance 80%.

I'm not very optimistic about future of our people. The growth is deepening and not widening. The top 10% is only growing.

800 million people in the country cannot even survive but for the monthly free food grains provided by the government.

UPA 2 and three terms of Modi has killed the aspirations and growth potential of the nation. Opportunity loss for one entire generation when youth population has been peaking.

We can only weep for India and Indians.

[Devdutt @notdevdutt]

@paulg To begin solving this problem, one might be better off rethinking the existing human traits as their AI equivalents instead:

Learning -> Pattern recognition
Creativity -> Combinatorial generation
Reasoning -> Probabilistic inference

[Naval @naval]

Nobody who’s actually good at making money needs to sell you a course on it.

[Devdutt @notdevdutt]

Nietzsche’s published writings don’t directly list “overwork, curiosity, and sympathy” as “three modern vices”. It seems like a loose paraphrase or modern restatement of his themes, not a quotation in works like The Gay Science or Beyond Good and Evil.

Nietzsche criticized the modern “busy” life and ceaseless work, but he didn’t present “overwork” as a codified “vice.” He also warned against shallow curiosity and praised a deeper, radical curiosity that questioned values. Nietzsche criticized “pity” (Mitleid) as life-denying, especially in works like The Antichrist and On the Genealogy of Morality. However, he didn’t discuss “sympathy for all” in those terms or label it a “modern vice” in a triad with overwork and curiosity.