News for hackers. Stuff that matters.

srbsa

Don't you remember the time that people were solving the 'Strait of Hormuz' crisis on LinkedIn with lines drawn on maps and showing their 'problem-solving' skills?

... or was it just my feed?

How much of code review friction comes from context outside the codebase?

I analysed 10K open-source PRs for something I'm building and found that anywhere between 42% and 67% of review friction is due to context beyond the codebase.From a workflow PoV, how annoying is the manual context synthesis to curate the prompt/agents.md? Is review a solved problem for teams with review agents, or is it now the real blocking step?

( Read More... | no comments | on HN )
Re:Parallelize yourself, not your agents on Thursday June 04, 2026 @01:48PM (#48402075)
Is it arguing that we need to supervise better than building systems that enable agents to make better decisions?

I would argue that unique decisions should only be made once (I'm building it). It is a gated mechanism, where confidence is 0<x<y<1 >y - decision delegated from x to y - rapid decision, upto 5 lines of context and simplified ask <x - full context for elaborate decision

Everything <y is fed back for making decisions. Plus, decisions are logged with fragments/context used.

By Amdahl's law, you'd have to automate 99.9% of decision-making to reach 1000x, 100x is at 99%. That's near impossible unless y is closer to 0, i.e., near unbound delegation.

Ask HN: Heavy coding-agent users what's your context plumbing? Where it fails?

From what I've seen it sits at three levels:Layer 1: Personal (memory MCP tools, personal *.md files) Layer 2: Repo (CLAUDE or AGENTS.md, in-repo docs) Layer 3: Org-wide (Unblocked, Glean, or similar)Most people have strong opinions on 1 and 2. Layer 3 seems underinvested.Curious, do you think 1 and 2 make 3 redundant, or would a strong 3 mean you'd stop/reduce patching with 1 and 2?

( Read More... | 2 comments | on HN )
I believe we have been there before; we responded with a similar emotional-visceral response during the printing press (replication of information), the industrial revolution (human labour), and the internet (distribution of information).

Now it's AI (cognition/human judgement); technological gains solve the bottlenecks. This is yet another layer. Amdahl's law.

I believe human societies will be a bell-curve distribution of all of them. The scary fraction will wield power/wealth; the indifferent will be the largest distribution, followed by the smallest fraction of benevolent people.

Re:Amdahl's Law for LLM generated code on Friday May 29, 2026 @12:12PM (#48325184)
Thank you! This changed how I was thinking about my product: a knowledge harness for coding agents. While I was discussing incidents & context-switching tax, the structural compounding is likely the real win.

Ask HN: Do coding agents need cross-tool org knowledge? Or, just good to have?

I've been talking to engineers, mostly in large teams. While they love cross-surface search with Glean, they still assimilate and curate the context for agents manually. It is especially useful during incidents and new onboarding.I've been building a knowledge substrate for engineering teams that is native to coding agents and pulls from all connected tools. It returns cross-source, fresh,...

( Read More... | no comments | 2 points | on HN )
The issue could be a matter of outrage, though, here in India the FOMO by parents (and students) is rather structural - been on since forever whether it is likes of Aptech/NIIT since late 90s.

Here are some pointers: High-school science grds that seek STEM undergrad education in India is over 5-million annually. Commerce & Humanities is looked down upon as barely has any value in terms career prospects (Some top high-schools don't even offer Humanities). For STEM subjects, over 5-million aspirants compete for less than 100K 'meaningful' seats - cuthroat yet?

FOMO in the failed education system is structural. Else to blame companies from byjus, to unacademy, to upgrad, all of which claim to deliver on "Career Changing Education" at high-markups, is lame!