Hacker News: carschno

New comment by carschno in "Google officially announces that ads will be included in AI Mode search results"

carschno — Thu, 21 May 2026 11:11:50 +0000

Don't _ads in search_ account for 100% of their search revenue? Does Google Search offer any other paid services?

New comment by carschno in "We stopped AI bot spam in our GitHub repo using Git's –author flag"

carschno — Tue, 19 May 2026 09:28:08 +0000

I don't know, and I think there is no easy answer. The point is: the investors don't know how to measure traction either, so they just measure GitHub activity instead, even at the very moment in which it becomes obvious that it does not capture actual traction. The absurdity lies in the statement that the developers still need to gain actual traction while putting additional effort into gaming that metric to satisfy their investors.

New comment by carschno in "We stopped AI bot spam in our GitHub repo using Git's –author flag"

carschno — Tue, 19 May 2026 04:47:45 +0000

> It's especially sensitive for a VC-backed startup that is measured thoroughly by GitHub activity, but we have to pull the trigger:

This sentence also illustrates the absurdity of this investment model. It imposes a trade-off between building good software, and complying with the investor's metrics. They probably call such metrics evidence-based, but this example shows that they arbitrarily capture some numbers to obscure the lack of meaningful measurements.

New comment by carschno in "VS Code inserting 'Co-Authored-by Copilot' into commits regardless of usage"

carschno — Sun, 03 May 2026 08:33:24 +0000

In literally must have missed that. When did Microsoft ever encourage energy saving? Is this related to power saving for extending laptop battery runtime? But then I don't get the link to renewable energy.

Anyway, I agree with the notion of the extreme energy-inefficiency of LLMs. The scale of it makes it hard to imagine any less efficient product will ever be invented.

New comment by carschno in "Google's 200M-parameter time-series foundation model with 16k context"

carschno — Tue, 31 Mar 2026 12:48:29 +0000

You could abstract speech or other audio as a series of sounds, where time is indeed a factor. Speech, however, has patterns that are more similar to written language than to seasonal patterns that are typically assumed in time series. While trained on different data, the architecture of TimesFM is actually similar to LLMs. But not identical, as pointed out at https://research.google/blog/a-decoder-only-foundation-model...:

> Firstly, we need a multilayer perceptron block with residual connections to convert a patch of time-series into a token that can be input to the transformer layers along with positional encodings (PE).

> [...]

> Secondly, at the other end, an output token from the stacked transformer can be used to predict a longer length of subsequent time-points than the input patch length, i.e., the output patch length can be larger than the input patch length.

New comment by carschno in "Personal Encyclopedias"

carschno — Thu, 26 Mar 2026 10:31:45 +0000

> On top of that, I exported my location timeline from Google Maps, my Uber trips, my bank transactions, and Shazam history. I would ask Claude Code to start with the photos and then gradually give it access to the different data exports.

Is anyone else feeling uncomfortable with that? It is a great project and I don't want to bash it with general concerns, but sharing all my financial and location details with any service seems like opening the floodgates to my house.

My concern is not even strictly related to AI, but about sharing all my most private data with any service. There is always a significant chance all of it is leaked sooner or later.

A Call for Meaningful Work at a Slower Pace

carschno — Thu, 05 Mar 2026 10:50:48 +0000

Article URL: https://jenteottenburghs.wordpress.com/2025/11/18/a-call-for-meaningful-work-at-a-slower-pace/

Comments URL: https://news.ycombinator.com/item?id=47260131

Points: 4

# Comments: 1

New comment by carschno in "My spicy take on vibe coding for PMs"

carschno — Wed, 04 Mar 2026 10:10:30 +0000

There are various technical corrections, with arguable pros and cons. However, they do not match the underlying problem stated above:

> the rise of business types in tech company leadership

EU MEPs let Chat Control fail

carschno — Tue, 03 Mar 2026 14:44:59 +0000

Article URL: https://www.heise.de/en/news/Setback-for-the-Commission-EU-MEPs-let-chat-control-fail-11197237.html

Comments URL: https://news.ycombinator.com/item?id=47233149

Points: 7

# Comments: 0

New comment by carschno in "Firefox 148 Launches with AI Kill Switch Feature and More Enhancements"

carschno — Tue, 24 Feb 2026 10:18:02 +0000

The page seems to be a copy from the original Mozilla press release from February 2nd: https://blog.mozilla.org/en/firefox/ai-controls/

It was discussed here: https://news.ycombinator.com/item?id=46858492

New comment by carschno in "Water 'Bankruptcy' Era Has Begun for Billions, Scientists Say"

carschno — Mon, 26 Jan 2026 13:52:51 +0000

Here is the paper on which the article is based: https://link.springer.com/article/10.1007/s11269-025-04484-0

New comment by carschno in "Anthropic's original take home assignment open sourced"

carschno — Wed, 21 Jan 2026 09:49:29 +0000

On the one hand, this exercise probably reflects a realistic task. Daily engineering work comprises a lot of reverse engineering and debugging of messy code. On the other hand, this does not seem very suitable as an isolated assignment. The lack of code base-specific context has a lot of potential for frustration. I wonder what they really tested on the candidates, and whether this was what they wanted to filter for.

New comment by carschno in "Investigating the methodological foundation of lesion network mapping"

carschno — Thu, 15 Jan 2026 14:37:21 +0000

Essentially, the LNM turns out to be unsuitable for various diagnostics it has been applied for:

> Our findings reveal a foundational limitation: at its core, LNM involves a repetitive sampling of one and the same FC matrix. As a result, it systematically maps sets of local brain changes—whether they are patient lesions, magnetic resonance imaging-derived alterations, synthetic or random—onto the same nonspecific properties of the used FC data, producing highly similar networks across conditions.

Investigating the methodological foundation of lesion network mapping

carschno — Thu, 15 Jan 2026 14:33:52 +0000

Article URL: https://www.nature.com/articles/s41593-025-02196-7

Comments URL: https://news.ycombinator.com/item?id=46633068

Points: 1

# Comments: 1

New comment by carschno in "Netflix Open Content"

carschno — Tue, 30 Dec 2025 10:33:52 +0000

The last addition was made in 2020.

New comment by carschno in "Are the Three Musketeers allergic to muskets? (2014)"

carschno — Wed, 10 Dec 2025 11:18:31 +0000

I suppose you are right about the history of firearms. However, the novel was written in 1844, more than 200 years after the time in which it is set. Which makes me wonder if the author (Alexandre Dumas) knew and cared about the historic facts.

New comment by carschno in "GitHub Actions has a package manager, and it might be the worst"

carschno — Mon, 08 Dec 2025 14:20:25 +0000

Good point, also to illustrate that open-source is not a panacea. It merely holds a higher potential for certain issues to be fixed/improved than.

New comment by carschno in "GitHub Actions has a package manager, and it might be the worst"

carschno — Mon, 08 Dec 2025 13:47:25 +0000

It is concerning that GitHub hosts the majority of open-source software, while actively locking its users into a platform that is based on closed source for eerything except Git itself. This issue with Actions shows how maintaining proprietary software inevitably ends up rather low on the priority list. Adding new features is much more marketable, just like for any other software product. Enshittification ensues.

For those who can still escape the lock-in, this is probably a good occasion to point to Forgejo, an open-source alternative that also has CI actions: https://forgejo.org/2023-02-27-forgejo-actions/ It is used by Codeberg: https://codeberg.org/

New comment by carschno in "DeepSeek OCR"

carschno — Mon, 20 Oct 2025 08:52:02 +0000

Technically not OCR, but HTR (hand-written text/transcript recognition) is still difficult. LLMs have increased accuracy, but their mistakes are very hard to identify because they just 'hallucinate' text they cannot digitize.

New comment by carschno in "Fastmail desktop app"

carschno — Mon, 13 Oct 2025 07:55:19 +0000

I think the actual question was: why would a mail provider develop their own an email client?