dbreunig's submissions

1.		A OSS Library with No Code, Only Specs (dbreunig.com)
		3 points by dbreunig 17 days ago \| past \| 2 comments
2.		2025 in Review: Jagged Intelligence Becomes a Fault Line (dbreunig.com)
		2 points by dbreunig 27 days ago \| past
3.		Why I write (and you should too) (dbreunig.com)
		1 point by dbreunig 30 days ago \| past
4.		Applied AI in 2025: From 'Naked' Model Calls to Tool Use Environment Calls (dbreunig.com)
		1 point by dbreunig 38 days ago \| past
5.		Enterprise Agents Have a Reliability Problem (dbreunig.com)
		2 points by dbreunig 50 days ago \| past
6.		Don't Fight the Weights: Learn to Spot Contexts That Go Against Training (dbreunig.com)
		3 points by dbreunig 75 days ago \| past
7.		About That MIT Report: Enterprise AI Looks Bleak, but Employee AI Looks Bright (dbreunig.com)
		3 points by dbreunig 4 months ago \| past
8.		AI Companies School Like Fish to the New Use Case (dbreunig.com)
		3 points by dbreunig 4 months ago \| past
9.		Monolingual speech recognition models beat multilingual models ~30x bigger (arxiv.org)
		3 points by dbreunig 4 months ago \| past
10.		Are Chatbots Incompatible with Advertising? (dbreunig.com)
		3 points by dbreunig 4 months ago \| past
11.		The AI Job Title Decoder Ring (dbreunig.com)
		91 points by dbreunig 5 months ago \| past \| 71 comments
12.		How will product management keep up with AI-accelerated coding? (dbreunig.com)
		3 points by dbreunig 5 months ago \| past
13.		GPT-OSS 20B delivers GPT-4.1 scores at 5x the speed (dbreunig.com)
		2 points by dbreunig 5 months ago \| past
14.		FLUX.1-Krea and the Rise of Opinionated Models (dbreunig.com)
		88 points by dbreunig 5 months ago \| past \| 34 comments
15.		Does the Bitter Lesson Have Limits? (dbreunig.com)
		180 points by dbreunig 5 months ago \| past \| 86 comments
16.		Moonshot Used RL for Qualitative Tasks to Write Better (dbreunig.com)
		2 points by dbreunig 5 months ago \| past
17.		Kimi 2 Became One of the Best Tool-Using Models (dbreunig.com)
		2 points by dbreunig 6 months ago \| past
18.		Rewriting Training Data Improved Kimi 2's Performance (dbreunig.com)
		1 point by dbreunig 6 months ago \| past
19.		Task Delegation by Experts Is the AI Metric That Matters (dbreunig.com)
		4 points by dbreunig 6 months ago \| past \| 1 comment
20.		Ask HN: How are startups dealing with key talent poaching?
		4 points by dbreunig 6 months ago \| past \| 2 comments
21.		Techniques for avoiding common long context fails (dbreunig.com)
		2 points by dbreunig 7 months ago \| past
22.		The Ways Long Contexts Fail (dbreunig.com)
		2 points by dbreunig 7 months ago \| past
23.		An Agentic Case Study: The Messy Work of Building a Gemini Agent to Play Pokémon (dbreunig.com)
		3 points by dbreunig 7 months ago \| past
24.		The drawbridges come up: the dream of a interconnected context ecosystem is over (dbreunig.com)
		111 points by dbreunig 7 months ago \| past \| 83 comments
25.		Let the Model Write the Prompt: Using DSPy to Decouple Your Task from the LLM (dbreunig.com)
		4 points by dbreunig 7 months ago \| past
26.		Comparing Claude System Prompts Reveal Anthropic's Priorities (dbreunig.com)
		135 points by dbreunig 7 months ago \| past \| 64 comments
27.		Using 'Slop Forensics' to Determine Model Lineage (dbreunig.com)
		3 points by dbreunig 8 months ago \| past
28.		Chatbots Trigger Guardrails Differently Based on Your Age, Gender, and Fandom (dbreunig.com)
		1 point by dbreunig 8 months ago \| past \| 1 comment
29.		Breaking Down Claude's 26k+ Token System Prompt (dbreunig.com)
		9 points by dbreunig 8 months ago \| past
30.		DuckDB is probably the most important geospatial software of the last decade (dbreunig.com)
		491 points by dbreunig 8 months ago \| past \| 165 comments
		More