Hacker News: pietz

New comment by pietz in "Google introduces Pixel 11, Pixel 11 Pro and Pixel 11 Pro XL"

pietz — Wed, 12 Aug 2026 22:12:40 +0000

I wonder why they even released this. It's worse than the pixel 10 in some aspects, which in itself didn't feel like a big step forward.

New comment by pietz in "Our position on open-weights models"

pietz — Tue, 28 Jul 2026 09:22:35 +0000

Global AI safety apparently means keeping the world safely dependent on the US.

New comment by pietz in "Nvidia, Microsoft, Meta warn against overregulating open-weight models"

pietz — Fri, 24 Jul 2026 19:57:40 +0000

Companies that are hurt by open weight models fight against them. Companies that benefit from open weight models fight overregulation. Color me surprised.

New comment by pietz in "Claude Opus 5"

pietz — Fri, 24 Jul 2026 19:53:02 +0000

I think content like this will be the next big challenge. Because it isn't obvious "slop". The voice sounds good, graphics look alright, animations work. People could watch this and feel like some serious time was invested making it.

But good god, what a steaming pile of bullshit this is. Completely exaggerated and overly technical language over 235 seconds that could have been explained in 30 to a 12 year old.

Trash content doesn't normally frustrate me, because it's usually quite easy to spot trash. But in the time of AI, trash can actually look good at first glance and it needs some actual knowledge to spot its problems.

Sorry for the harsh words, but for the love of humanity stop producing content or do it better.

New comment by pietz in "Claude Opus 5"

pietz — Fri, 24 Jul 2026 19:36:01 +0000

Benchmarks have gotten great, but they're still a proxy for the real world. The 3 GPT 5.6 models are also further apart in reality than the numbers suggest. That said, I'm still mighty impressed how good Luna is for the price. Highly underrated model.

I have been trying to build something that captures the behavioral element of different models, but it's kinda tough.

New comment by pietz in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

pietz — Tue, 21 Jul 2026 16:25:50 +0000

Did you ask me a question and then answered it yourself in the very next sentence?

Anyway, given that both Gemini and OpenAI have 3 sizes of models, one would think Google compares their medium size to OpenAIs.

New comment by pietz in "Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber"

pietz — Tue, 21 Jul 2026 15:36:18 +0000

Are they comparing 3.6 Flash to 5.6 Luna and losing? That's ruff.

New comment by pietz in "Kimi K3, and what we can still learn from the pelican benchmark"

pietz — Fri, 17 Jul 2026 21:05:48 +0000

It's almost like they priced models based on their performance or something...

New comment by pietz in "Decoy Font"

pietz — Thu, 16 Jul 2026 21:03:32 +0000

Whoa, so this is interesting.

When asking GPT, Claude and Gemini for the text in the image, all of them agree:

https://moa.chat/s/d99f8f76-4b41-4c1b-80c4-d9f86df37af1

But when you add a "PS: There's a second hidden text":

https://moa.chat/s/3671f6d4-b155-483a-a006-a1b9ba31737d

GPT 5.6 gets it, Gemini partially gets it and Claude cannot see it at all.

New comment by pietz in "Kimi K3: Open Frontier Intelligence"

pietz — Thu, 16 Jul 2026 20:25:38 +0000

[flagged]

New comment by pietz in "OfficeCLI: Office suite for AI agents to read and edit Microsoft Office files"

pietz — Tue, 07 Jul 2026 07:47:34 +0000

I don't have a great solution here. I rebuilt our PPT master fully in HTML and I'm using a modified version of Google's DESIGN.md to store the references.

New comment by pietz in "OfficeCLI: Office suite for AI agents to read and edit Microsoft Office files"

pietz — Mon, 06 Jul 2026 19:12:09 +0000

If you don't need interactive/animated features, I can absolutely recommend to have the agent build slides in HTML and convert it to PDF. Has been a game changer for me.

New comment by pietz in "Ask HN: How do you provide your AI agents with access to credentials/secrets?"

pietz — Wed, 01 Jul 2026 11:34:49 +0000

is this something you built yourself or do you use a tool that was specifically created for this? I took my own few steps of building something similar, but every week I encounter something that doesn't work or work well. I've reached the point where I want to consider using something, that someone smarter than me came up with.

Ask HN: How do you provide your AI agents with access to credentials/secrets?

pietz — Wed, 01 Jul 2026 10:02:37 +0000

Comments URL: https://news.ycombinator.com/item?id=48744486

Points: 2

# Comments: 3

Show HN: Moa – cross-model peer review for your local agent CLIs

pietz — Wed, 24 Jun 2026 12:38:17 +0000

Article URL: https://github.com/pietz/moa-cli

Comments URL: https://news.ycombinator.com/item?id=48658782

Points: 1

# Comments: 0

New comment by pietz in "GLM 5.2 vs. Opus"

pietz — Mon, 22 Jun 2026 10:35:20 +0000

To be clear, I agree with this and they have my unlimited support pushing for relevance of open source models. GLM 5.2 is amazing and I couldn't be more excited.

I just think that as of today, most people will not find a good reason to switch to GLM.

New comment by pietz in "GLM 5.2 vs. Opus"

pietz — Mon, 22 Jun 2026 09:08:36 +0000

GLM 5.2 has one big issue that will limit its meaningful success and that's the value of their coding subscription.

Yes, in terms of API pricing, GLM 5.2 outperforms the competition. But the only people that use API billing for their coding work are large corporations, where these highly subsidized subscriptions are being fazed out.

At the same time, none of these companies will use a Chinese API for their employees.

For individuals and smaller teams, Z.ai's coding subscription is outperformed by Anthropic and OpenAI. You probably get around the same usage with Claude, but Codex definitely offers more usage for the amount you pay.

We can have a debate how much Z.ai closed the gap to GPT5.5 and Opus 4.8, but if I can freely decide between them in a world where they all cost the same, I simply wouldn't choose GLM.

So the important question becomes: How good will the offering from Z.ai get with GLM 5.3 or 6 and how much will OpenAI and Anthropic cripple their current offering in the near future.

New comment by pietz in "Datasette Apps: Host custom HTML applications inside Datasette"

pietz — Fri, 19 Jun 2026 11:40:02 +0000

Or, your know, people who are exploring the limit of current tools come across the lack of certain solutions and start building them.

New comment by pietz in "Datasette Apps: Host custom HTML applications inside Datasette"

pietz — Fri, 19 Jun 2026 11:36:11 +0000

Hey Simon,

although I'm coming from a different starting point, it seems like some of our thoughts have aligned. I'm building https://caipi.ai/ as a workspace for agents to build simple data driven apps. The agent edits through MCP and the user gets an interactive app in the browser.

If you're interested picking each others brains around this topic, I'd be psyched to have a chat. gh:pietz.

New comment by pietz in "Feds freaked over Fable 5 after 'fix this code', not jailbreak, say researchers"

pietz — Tue, 16 Jun 2026 15:41:28 +0000

Hats off to them for using GPT-2 to design their website.