tirumaraiselvan

> Tier 3 incidents: Novel, complex, cascading failures stay human-led, with AI providing supporting context.

> But here’s the catch.

> If human SREs (okay, now we have to use adjectives like human ) only engage with Tier 3 incidents, i.e. the novel, never-before-seen outages, where do they build the intuition to handle them?

How did they do this before AI? I have done a bunch of complex production level debugging and these Tier 3 incidents are always "starting from scratch" never seen before issues. The main "intuition" I built is persistence and verifying/refuting random hypothesis. In other words, I think we can build an intuition for Tier 3 issues only by engaging with Tier 3 issues.

Although, psychologically, some success in T1 and T2 issues is what gives us the confidence to even approach the T3 issues.

The fact that this question is getting a due regard now, then before AI, shows that all we care about from education is getting a job that pays. And we are worried that jobs are being taken away by AI.

Otherwise this question should bother a parent in every generation.

How is it industrial revolution era theory anymore? I learn pretty much everything Languages, Science, Maths, Arts etc

Maybe mass education picked up during industrial revolution because of the immense needs of operating factories but the curriculum has obviously been reformed in every generation.

It also leaves the part of reviewing the code (which I have realised is really quite different skillset from writing code - it needs a sense of something is fishy here or why is this even here or can't this look better).

It may be obvious to conclude that good programmers are good reviewers but I think all variations exist. I wonder how many people don't like their jobs anymore because they don't like reviewing code written by AI.

I think summary is:

1. Paper has not been replaced by any tech because its form still has unique use-cases

2. AI is stuck in "chat" form and not all use-cases can be solved with chat.

Hence "AI will never replace paper" ..... well anything can happen really isn't it? 1 and 2 are quite independent statements.

Out of curiosity, which tool do you recommend?

I am the OP. I added some text while submitting so people know who the author is. Not AI generated.

But the real irony is that even the author points that their authentic work is under constant suspicion because "what if this is AI" has prevaded everyone's consciousness.

Brilliant piece by Booker Prize winning author, Arundhati Roy. She manages to lay bare the entire global politics of AI, starting with the seemingly inconsequentual em-dash.

Response from the organizers: https://proofsandprompts.com/2026/09/10/a-response-from-math...

Now we know why OpenAI told its investors to avoid Glean[1]

[1]: https://www.reuters.com/technology/openai-tells-investor-not...

This article is getting a lot of hate but honestly it does have good amount of useful content learned through practical experience, although at an abstract level. For example, this section:

``` The teams that succeed don’t just throw SQL schemas at the model. They build:

Business glossaries and term mappings

Query templates with constraints

Validation layers that catch semantic errors before execution ```

Unfortunately, the mixing of fluffy tone and high level ideas is bound to be detested by hands on practitioners.

A simple way is perhaps implement a text-to-metrics system where metrics could be defined as SQL functions.

Exactly! Midnights Children and Moors Last Sigh are so non-linear that one cannot expect to get the hang of it till they are atleast 50 pages through. It's usually on the second reading that the amazingness of those initial pages is felt.

Big Rushdie fan here. I used to think it could be because of a lack of cultural context especially for books like Midnights Children and Moors Last Sigh but now I also think that it could also be a matter of taste. Rushdie himself quotes Milan Kundera who said: "...that the novel descended from two parents, Samuel Richardson’s Clarissa and Laurence Sterne’s Tristram Shandy". The latter is a style of writing where all conventional rules of literature are broken, it's just wildly creative so to speak. Rushdie's Midnights Children and James Joyce's Ulysses fall into this category.

If you want to see the other style of writing in Rushdie, I can suggest Shalimar The Clown or The Ground Beneath Her Feet. But these are nowhere near as grand as Midnights Children.

In either category, a fair amount of interest in history helps to enjoy his books.

Hi HN

PromptQL is a programmatic approach to LLM tool composition which has shown remarkable improvement in accuracy (~2x) and repeatability (~4x) when compared to traditional tool calling and chaining approaches.

Fundamentally, PromptQL separates the creation of a query plan from the execution of the query plan. PromptQL uses the LLM to first create a program to compose the different tool calls hence avoiding the challenges with in-context tool chaining like accuracy, repeatability and context limits.

It was published in New Yorker https://www.newyorker.com/magazine/1946/08/31/hiroshima

Fun fact the cover image if this edition was kind of a decoy (perhaps to accentuate the shock): https://www.newyorker.com/magazine/1946/08/31

> RAG doesn't have to involve vector search.

This. Not sure why RAG triggers vector search for everyone. Retrieval Augmented Generation is as generic as it can get.

Why do you think that's more reasonable :)

To me it seems a more risky use-case since you don't have control/observability over what an untrusted user is asking for?