Hacker News: stygiansonic

New comment by stygiansonic in "Every AI model has the same "controversial" take on humans"

stygiansonic — Fri, 23 Oct 2026 01:40:04 +0000

Gettier actually criticized justified true belief (JTB) as insufficient for knowledge and provided counter examples known as Gettier problems

New comment by stygiansonic in "Every AI model has the same "controversial" take on humans"

stygiansonic — Fri, 23 Oct 2026 01:37:35 +0000

Indeed. In fact this criticism of JTB is known as the Gettier problem: https://iep.utm.edu/gettier/

New comment by stygiansonic in "Dogs seem to agree with humans that 'bouba' is round and 'kiki' is spiky"

stygiansonic — Mon, 19 Oct 2026 16:00:15 +0000

Related to the Bouba/Kiki effect is the Molyneux problem: https://en.wikipedia.org/wiki/Molyneux%27s_problem

The experimental evidence there, though limited, suggested no innate connection between touch and sight.

Anthropic Model Submitted a False Homicide Tip to Philadelphia Police

stygiansonic — Sat, 10 Oct 2026 00:52:49 +0000

Article URL: https://www.engadget.com/2282713/an-anthropic-model-submitted-a-false-homicide-tip-to-philadelphia-police/

Comments URL: https://news.ycombinator.com/item?id=50028365

Points: 5

# Comments: 1

New comment by stygiansonic in "Sakana Fugu"

stygiansonic — Mon, 22 Jun 2026 03:21:29 +0000

From a brief reading of what Fusion does: https://openrouter.ai/docs/guides/features/plugins/fusion

Looks like Fusion calls a bunch of models and then uses an LLM to synthesize the results, and pass to another model for final output.

Fugu looks like it's doing something different? Using an LLM earlier on in the flow as an orchestrator to decide which other LLMs to call. More coordinator than simply synthesizing results, and more "agentic".

It's interesting because it's all exposed behind a single OpenAI compatible endpoint (Responses API?) and so then presumably someone could use this for one of their single agents. Now you have agent-of-agents, nested in some sense. The token usage increases accordingly!

New comment by stygiansonic in "Iran strikes leave Amazon availability zones "hard down" in Bahrain and Dubai"

stygiansonic — Fri, 03 Apr 2026 22:46:03 +0000

(Perpetrators also not caught)

Family of Tumbler Ridge shooting victim suing OpenAI

stygiansonic — Tue, 10 Mar 2026 01:15:07 +0000

Article URL: https://www.cbc.ca/news/canada/british-columbia/openai-sued-tumbler-ridge-victim-9.7121635

Comments URL: https://news.ycombinator.com/item?id=47318006

Points: 9

# Comments: 0

New comment by stygiansonic in "Google Engineer Found Guilty of Sending AI Secrets to China"

stygiansonic — Fri, 30 Jan 2026 23:19:29 +0000

  The jury found that Ding stole trade secrets relating to the hardware infrastructure and software platforms that allow Google’s supercomputing data center to train and serve large AI models. The trade secrets contained detailed information about the architecture and functionality of Google’s custom Tensor Processing Unit chips and systems and Google’s Graphics Processing Unit systems, the software that allows the chips to communicate and execute tasks, and the software that orchestrates thousands of chips into a supercomputer capable of training and executing cutting-edge AI workloads. The trade secrets also pertained to Google’s custom-designed SmartNIC, a type of network interface card used to facilitate high speed communication within Google’s AI supercomputers and cloud networking products.

Apple picks Gemini to power Siri

stygiansonic — Mon, 12 Jan 2026 15:22:21 +0000

Article URL: https://www.cnbc.com/2026/01/12/apple-google-ai-siri-gemini.html

Comments URL: https://news.ycombinator.com/item?id=46589675

Points: 1038

# Comments: 651

New comment by stygiansonic in "Sampling at negative temperature"

stygiansonic — Sun, 11 Jan 2026 21:06:02 +0000

Neat experiment that gives a mechanistic interpretation of temperature. I liked the reference to the "anomalous" tokens being near the centroid, and thus having very little "meaning" to the LLM.

US authorities shut down major China-linked GPU smuggling operation

stygiansonic — Tue, 09 Dec 2025 13:41:36 +0000

Article URL: https://www.justice.gov/usao-sdtx/pr/us-authorities-shut-down-major-china-linked-ai-tech-smuggling-network

Comments URL: https://news.ycombinator.com/item?id=46204829

Points: 2

# Comments: 0

Anthropic to buy $30B in Azure capacity in deal with Microsoft, Nvidia

stygiansonic — Tue, 18 Nov 2025 15:05:33 +0000

Article URL: https://www.cnbc.com/2025/11/18/anthropic-ai-azure-microsoft-nvidia.html

Comments URL: https://news.ycombinator.com/item?id=45967145

Points: 3

# Comments: 1

New comment by stygiansonic in "We bought the whole GPU, so we're damn well going to use the whole GPU"

stygiansonic — Thu, 02 Oct 2025 16:11:47 +0000

That paper doesn’t seem to be about security vulnerabilities in MiG but rather using it to improve workload efficiency

New comment by stygiansonic in "Pentagon Pizza Index"

stygiansonic — Thu, 31 Jul 2025 01:59:27 +0000

Wonder why they haven’t gotten an in house pizzeria yet to reduce the signal on this side channel leak

New comment by stygiansonic in "Gemma 3n preview: Mobile-first AI"

stygiansonic — Tue, 20 May 2025 19:33:57 +0000

From the article it appears to be something they invented:

> Gemma 3n leverages a Google DeepMind innovation called Per-Layer Embeddings (PLE) that delivers a significant reduction in RAM usage.

Like you I’m also interested in the architectural details. We can speculate but we’ll probably need to wait for some sort of paper to get the details.

New comment by stygiansonic in "Reservoir Sampling"

stygiansonic — Fri, 09 May 2025 01:15:52 +0000

Interesting! If Knuth is not the original author then they’ve been lost to the sands of time

New comment by stygiansonic in "Reservoir Sampling"

stygiansonic — Thu, 08 May 2025 17:43:44 +0000

Great article and nice explanation. I believe this describes “Algorithm R” in this paper from Vitter, who was probably the first to describe it: https://www.cs.umd.edu/~samir/498/vitter.pdf

New comment by stygiansonic in "The Burnout Machine"

stygiansonic — Thu, 20 Mar 2025 20:14:34 +0000

The article mentions this union, not sure if it meets your definition of success: https://www.alphabetworkersunion.org/our-wins

New comment by stygiansonic in "Superintelligence Strategy"

stygiansonic — Wed, 05 Mar 2025 21:28:35 +0000

When subtlety proves too constraining, competitors may escalate to overt cyberattacks, targeting datacenter chip-cooling systems or nearby power plants in a way that directly—if visibly—disrupts development. Should these measures falter, some leaders may contemplate kinetic attacks on datacenters, arguing that allowing one actor to risk dominating or destroying the world are graver dangers, though kinetic attacks are likely unnecessary. Finally, under dire circumstances, states may resort to broader hostilities by climbing up existing escalation ladders or threatening non-AI assets. We refer to attacks against rival AI projects as "maiming attacks."

Superintelligence Strategy

stygiansonic — Wed, 05 Mar 2025 21:19:37 +0000

Article URL: https://www.nationalsecurity.ai/

Comments URL: https://news.ycombinator.com/item?id=43272607

Points: 33

# Comments: 62