Submissions from quesma.com

		We hid backdoors in ~40MB binaries and asked AI + Ghidra to find them (quesma.com)
		196 points by jakozaur 8 hours ago \| past \| 88 comments
		We hid backdoors in binaries – Opus 4.6 found 49% of them (quesma.com)
		2 points by stared 7 days ago \| past \| 1 comment
		BinaryAudit: Can AI find backdoors in raw machine code? (quesma.com)
		2 points by stared 9 days ago \| past \| 1 comment
		We hid backdoors in binaries – Opus 4.6 found 49% of them (quesma.com)
		5 points by stared 11 days ago \| past \| 2 comments
		We hid backdoors in binaries – Opus 4.6 found 49% of them (quesma.com)
		6 points by stared 12 days ago \| past \| 2 comments
		Reverse Engineering River Raid with Claude, Ghidra, and MCP (quesma.com)
		3 points by stared 20 days ago \| past
		Benchmarking OpenTelemetry: Can AI trace your failed login? (quesma.com)
		144 points by stared 24 days ago \| past \| 80 comments
		OTelBench: Can AI instrument OpenTelemetry? (quesma.com)
		5 points by jakozaur 27 days ago \| past \| 1 comment
		Reverse Engineering River Raid with Claude, Ghidra, and MCP (quesma.com)
		3 points by stared 29 days ago \| past
		Benchmarking OpenTelemetry: Can AI trace your failed login? (quesma.com)
		4 points by stared 33 days ago \| past
		Vibe coding needs Git blame (quesma.com)
		1 point by stared 43 days ago \| past
		Prompts are (not) the new source code (quesma.com)
		6 points by stared 44 days ago \| past \| 1 comment
		2025 took AI from party tricks to production tools (quesma.com)
		2 points by stared 50 days ago \| past \| 1 comment
		Migrating CompileBench to Harbor: standardizing AI agent evals (quesma.com)
		6 points by stared 62 days ago \| past
		Antigravity feels heavy and Claude Skills are light (quesma.com)
		1 point by stared 68 days ago \| past
		Outside of the bubble, AI is Black Mirror (quesma.com)
		7 points by stared 79 days ago \| past
		Nano Banana Pro: raw intelligence with tool use (quesma.com)
		4 points by jakozaur 87 days ago \| past
		Nano Banana Pro: raw intelligence with tool use (quesma.com)
		3 points by amrrs 89 days ago \| past
		Nano Banana Pro: raw intelligence with tool use (quesma.com)
		5 points by stared 3 months ago \| past \| 2 comments
		AI for coding is still playing Go, not StarCraft (quesma.com)
		1 point by stared 3 months ago \| past
		A postmortem on our $2.5M database gateway: lessons from pilot purgatory (quesma.com)
		6 points by jakozaur 3 months ago \| past
		We raised $2.5M for a database gateway and failed: Postmortem (quesma.com)
		3 points by groseje 3 months ago \| past
		AI for coding is still playing Go, not StarCraft (quesma.com)
		1 point by stared 4 months ago \| past
		The security paradox of local LLMs (quesma.com)
		160 points by jakozaur 4 months ago \| past \| 87 comments
		Local LLMs are worse for security (quesma.com)
		1 point by jakozaur 4 months ago \| past
		AI for coding is still playing Go, not StarCraft (quesma.com)
		5 points by stared 4 months ago \| past \| 2 comments
		AI for coding is still playing Go, not StarCraft (quesma.com)
		1 point by aamederen 4 months ago \| past \| 1 comment
		Observability in Go: What Real Engineers Are Saying in 2025 (quesma.com)
		1 point by darccio 4 months ago \| past
		CompileBench: Can AI Compile 22-year-old Code? (quesma.com)
		148 points by jakozaur 5 months ago \| past \| 65 comments
		Tau² benchmark: How a prompt rewrite boosted GPT-5-mini by 22% (quesma.com)
		197 points by blndrt 5 months ago \| past \| 65 comments
		More