Hacker News: ironSkillet

New comment by ironSkillet in "Harnessing the Universal Geometry of Embeddings"

ironSkillet — Mon, 07 Sep 2026 00:44:48 +0000

I am not familiar with the standards of publishing in machine learning, but as someone trained in a mathematics background, this paper seems relatively light on details and heavy on exposition. Is that typical? Is this a really novel idea? Not trying to be snarky, just trying to understand how meaningful this is.

New comment by ironSkillet in "Geometric Reasoning"

ironSkillet — Sat, 15 Aug 2026 13:57:54 +0000

The entire thing is so generic and full of wise sounding things with no deep concrete content. It screams "an AI wrote this and a human thought it looked ok and pushed it".

New comment by ironSkillet in "Financing the AI boom: from cash flows to debt [pdf]"

ironSkillet — Wed, 15 Jul 2026 16:27:44 +0000

I think we can agree to disagree here. I don't see how a company needs to have an $8000 increase in revenue to justify a $240 software purchase. You are assuming that the current operating cost for every dollar of revenue is also applied to every incremental dollar in revenue gained from software efficiency, and that is just not true.

New comment by ironSkillet in "Financing the AI boom: from cash flows to debt [pdf]"

ironSkillet — Wed, 15 Jul 2026 11:25:45 +0000

I understand that. The AI software is meant to be a productivity enhancer for the employees using it. Other than the licenses, some training etc, there are no operating costs associated with it. Just by using the software, I don't suddenly have to pay more for salaries, retirement plans, etc, which are things that in aggregate produce the 3% margin. Maybe I have to pay more in logistics because I'm moving more product now, but I think the point stands.

New comment by ironSkillet in "Financing the AI boom: from cash flows to debt [pdf]"

ironSkillet — Wed, 15 Jul 2026 02:49:02 +0000

I'm not an MBA over here, but this math seems wrong. If they are spending $240 in increased costs, then they only have to make about $247 in additional revenue from that spend to preserve a 3% margin. That seems much more reasonable if it increases the probability that customers find the product they are looking for and have a good experience.

New comment by ironSkillet in "Meta tells staff it will cut 10% of jobs"

ironSkillet — Fri, 24 Apr 2026 02:31:57 +0000

I interviewed someone recently who worked at Meta a couple years ago. He was a software engineer, was paid a bunch of money to mostly up dashboards all day, and eventually quit because it was neither interesting nor challenging.

New comment by ironSkillet in "Fundamental Theorem of Calculus"

ironSkillet — Thu, 23 Apr 2026 11:19:41 +0000

No that's not true either. A quick Google will reveal many examples, in particular the "Cantor set".

New comment by ironSkillet in "Fundamental Theorem of Calculus"

ironSkillet — Thu, 23 Apr 2026 11:16:47 +0000

It is not determined by the derivative, it's the antiderivative, as someone else mentioned. The derivative is the rate of change of a function. The "area under a curve" of the graph of a function measures how much the function is "accumulating", which is intuitively a sum of rates of change (taken to an infinitesimal limit).

New comment by ironSkillet in "There is No Quintic Formula [video]"

ironSkillet — Sun, 30 Nov 2025 23:33:36 +0000

This is some high quality content. Love the visual animations to go along with the mathematical ideas. Did a great job helping to tie the algebra to geometric intuition, but I think the importance of commutators could have gotten a little bit more exposition.

New comment by ironSkillet in "A triangle whose interior angles sum to zero"

ironSkillet — Sat, 29 Nov 2025 00:55:32 +0000

The disk model of hyberolic geometry is made to map hyperbolic 2 space (which is infinite in area) into the finite interior of the disk. In order to capture this, the normal euclidean notion of distance is distorted by a function which allows "distances" to go to infinity as a curve approaches the boundary of the disk.

New comment by ironSkillet in "I mathematically proved the best "Guess Who?" strategy [video]"

ironSkillet — Sat, 29 Nov 2025 00:48:10 +0000

Binary search minimizes the number of expected moves until you find the target. If you are already ahead, this is a natural thing to want to do. The reason why this doesn't work when you're behind is that your opponent can also do that and probabilistically maintain their lead.

New comment by ironSkillet in "NY school phone ban has made lunch loud again"

ironSkillet — Thu, 06 Nov 2025 00:14:44 +0000

I think the user means that bullies in school face little consequences, but a bully at work may get called by HR and potentially disciplined.

New comment by ironSkillet in "AI gets more 'meh' as you get to know it better"

ironSkillet — Thu, 09 Oct 2025 10:44:36 +0000

In my mind that is a problem with your lazy developer colleague, not AI as a whole. You can't expect it to be right on the first try (just like human code), you have to iterate with it and have the experience to know when it's off track and you have to take over.

New comment by ironSkillet in "AI gets more 'meh' as you get to know it better"

ironSkillet — Wed, 08 Oct 2025 23:36:57 +0000

I don't know about other use cases, but AI is definitively a game changer for software development. You still need to know what you're doing and test/think critically about what it's giving you, but the body of software problems that you can conceptually treat as "boilerplate" becomes massively larger with the help of a good AI coding tool.

New comment by ironSkillet in "Wikipedia survives while the rest of the internet breaks"

ironSkillet — Fri, 05 Sep 2025 02:22:06 +0000

It seems like there is a universal sense in which statements like 1+1=2" or "7 is a prime number" are true, no?

New comment by ironSkillet in "Putin's nuclear files"

ironSkillet — Wed, 28 May 2025 21:40:03 +0000

Almost certainly a Western intelligence operation to send a message to Putin.

New comment by ironSkillet in "The Web Is Broken – Botnet Part 2"

ironSkillet — Sun, 20 Apr 2025 00:59:04 +0000

I have found that the dependency injection pattern makes it far easier to write clean tests for my code.

New comment by ironSkillet in "AI models miss disease in Black and female patients"

ironSkillet — Fri, 28 Mar 2025 00:24:57 +0000

You and the article are both correct. The disease does present itself differently as a function of these other characteristics, so since the training dataset doesn't contain enough samples of these different presentations, it is unable to effectively diagnose.

New comment by ironSkillet in "'Next-Level' Chaos Traces the True Limit of Predictability"

ironSkillet — Sat, 08 Mar 2025 00:34:20 +0000

This is a popular science article, not a scientific paper.

New comment by ironSkillet in "The CAP theorem of Clustering: Why Every Algorithm Must Sacrifice Something"

ironSkillet — Fri, 27 Dec 2024 01:10:34 +0000

The "richness" definition also seemed hand wavey to me so I looked at the referenced paper. The actual definition of "richness" of an algorithm is that for any arbitrary partition P of your original data (singletons, one cluster, etc), there is a distance function on the data, which when used in the clustering algorithm, produces P.