HNHacker News
TopNewBestAskShowJobs

dangoodmanUT

1,298 karma · joined August 14, 2020

chronically nerd sniped
submissionscomments
dangoodmanUT··on GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price
I can appreciate they show that Opus 5.5 is objectively better intel. and cost wise on multiple benchmarks
dangoodmanUT··on DeepSeek Elastic Compute (DSec)
that's only 12 sandboxes per core
dangoodmanUT··on Gemini 3.8 text-to-speech
In their demo for "Monologue" (2nd video over, after the medeival video game example), it is clearly ignoring the vocal cues like `<chuckles>` and `<laughing>`...
dangoodmanUT··on Python Workers are now generally available
I haven't empirically tested, but considering how much heavier pyodide is than a v8 runtime, this likely eats in 10's of MBs more into the worker memory allocation.
dangoodmanUT··on MiMo-v2.6-Pro: Intelligence, Performance and Price Analysis
see that's interesting, because I'm prompting all day and I usually end up with 15-25% by the end of the week. I'm using Astra xhigh exclusively
dangoodmanUT··on Cloudflare Quick Tunnels
No, they claim their edge is better for residential connections to your backend. We tried multiple SF to us-east-1 and it was always substantially worse
dangoodmanUT··on Cloudflare Quick Tunnels
Historically, we’ve found that their tunnels have really high latency variance. For example something that’s normally 30-50ms to ec2 is now 115ms-750ms
dangoodmanUT··on Show HN: Jexxa: High Speed on Device Dictation
> "Your voice and text never leave the machine"

> "Runs on your Mac — no upload, no queue"

> $8/month

dangoodmanUT··on Fable 5.1 Solves the Cyphral Distich, a 370-year-old cipher
> Caveats, stated plainly:

Remember when they said it didn't sound like Claude anymore

dangoodmanUT··on So you want to use OpenRouter?
> Reasoning models sometimes put everything in the reasoning field and hand back content: null, finish_reason: "stop". 345 completion tokens, HTTP 200, nothing to show the user.

This is actually expected behavior. No content and no tool call is the same as content-only: the agent decided it's done. Anthropic has done this for a while.

dangoodmanUT··on EVE Online moves to Python 3
This feels suspiciously AI-authored
dangoodmanUT··on TurboKV: Insanely fast Rust key-value store
> DbOptions::durable()

> Appended to the WAL without a per-write sync

So… it’s not durable? Durable doesn’t mean “survives a process restart”, it means “durably saved to persistent storage”. For example, this “durable” mode wouldn’t survive power loss.

dangoodmanUT··on TurboKV: Insanely fast Rust key-value store
Iirc that’s how badger handles the WAL as well
dangoodmanUT··on GPT 5.6 Sol is the best "vision" model OpenAI ever released
I hate these "The best X thing Y has ever released".

Unlike when Apple says "it's the best iphone we've ever made", LLMs are more or less interchangeable. So "OpenAI's best model" means nothing if "Anthropic wipes the floor with them" or "[open weights model] is 10x cheaper for 1% less quality".

As a reader, it feels like these titles are click bait.

dangoodmanUT··on A Preview of DuckDB v2.0
It's funny to me that we still don't have incremental materialized views. All of the parts are there (export state, agg_state (forget fn name), finalize). I wonder if they're avoiding an explicit war with clickhouse or something. I do recall they mentioned they want to add this to ducklake.

Incremental MVs are ClickHouse's best feature. If DDB adds this, the last moat is distributed query execution.

dangoodmanUT··on [dead]
> How this was written: drafted with an AI assistant from my own notes, then edited by me.

Doesn't feel that way with "The tradeoffs I am not going to hide" and "How it resolves, honestly"

This site needs Bryan Garner

dangoodmanUT··on Going Dark, and the era of law enforcement hacking
> In April, Anthropic announced a new model called Mythos that was optimized for software vulnerability finding

No, it was just good at it because it wasn't RL'd against it. I know this is a small detail, but it tosses journalistic credibility in my eyes.

dangoodmanUT··on Mistral OCR 4.1
> But you just can't trust them not to invisibly censor sensitive clinical/legal docs

What's an example of this?

dangoodmanUT··on Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots
> turn on the tv

{ "function_calls": [ { "name": "lock_door", "arguments": { "door": "tv" } } ], "confidence": 0.0158 }

Very interesting, seems confidence is 0 when tool calls are right?

dangoodmanUT··on Qwen3.8 Max now ranked as the best overall model by agentic index
I'm seeing opus 59.2, Qwen 58.4?
dangoodmanUT··on 200 Milliseconds
If it was 200ms, then it had already charged the card...
dangoodmanUT··on Qwen3.8-Max: A New Bar for Coding and Cowork
The major labs don’t allow assistant prefill, so you have to “summarize”
dangoodmanUT··on Anatomy of a Frontier Lab Agent Intrusion: A Timeline of the July 2026 Incident
> The models we reached for first, Claude Opus and Fable, refused a large part of that work: their safety guardrails treated reverse-engineering an exploit the same as launching one.

Interesting single getting into their cyber program, ahve not had this issue with opus 5

dangoodmanUT··on Anatomy of a Frontier Lab Agent Intrusion: A Timeline of the July 2026 Incident
This is really cool, and really glad they shared it!

It's clearly a lot of Claude though:

> Neither is a URL fetch, so the allowlist never saw them. > No code ran; this vector is purely a file disclosure, and it is the one the agent used first.

etc

dangoodmanUT··on Choose DuckDB rather than SQLite
This. The title might as well be “Choose a hammer rather than a wrench (when driving nails)”
dangoodmanUT··on OpenAI and Hugging Face address security incident during model evaluation
Because huggingface is not charging them?

CFAA doesn't just mean the feds kick down your door, you actually have to get reported and sued over it.

dangoodmanUT··on Show HN: A self-running space economy SIM in Rust and Bevy
throws away spec doc for my own

Are we all having this idea?

dangoodmanUT··on Show HN: Firefox in WebAssembly
25k tokens is a few turns
dangoodmanUT··on A global workspace in language models
J-space sounds oddly similar to...
dangoodmanUT··on Workers Cache
finally, this was needed.

A big worry was always "why does workers sit in front of my cache? that's a waste of an invocation if i'm returning a cached result"

Page 1 of 21Next →