HNHacker News
TopNewBestAskShowJobs

drakenot

1,332 karma · joined February 14, 2015

submissionscomments
drakenot··on Ccusage: A CLI tool for analyzing Claude Code usage from local JSONL files
Have CC output a plan.md file.
drakenot··on Colombia seizes first unmanned narco-submarine with Starlink antenna
Remote tracking?
drakenot··on OpenAI releases image generation in the API
Does the AI have the same content restrictions that the chat service does?
drakenot··on Show HN: I open-sourced my AI toy company that runs on ESP32 and OpenAI realtime
Yes, I mentioned this in the comment.

I think it is closer, although still even it has a cold start problem. Once you are connected and in-session, it is a better experience.

There is still some "turn based" conversational aspect to it that can be awkward but it is much better. It also helps that you can "tap and hold" to override, which is a bit of a hack but works well in practice for that mobile use-case.

drakenot··on Show HN: I open-sourced my AI toy company that runs on ESP32 and OpenAI realtime
Something that really kills the 'effect' of most of the Voice > AI demos that I see is the cold start / latency.

The OpenAI "Voice Mode" is closer, but when we can have near instantaneous and natural back and forth voice mode, that will be a big in terms of it feeling magical. Today, it is say something, awkwardly wait N seconds then listen to the reply and sometimes awkwardly interrupt it.

Even if the models were no smarter than they are today, if we could crack that "conversational" piece and performance piece, it would be a big difference in my opinion.

drakenot··on JetBrains IDEs Go AI: Coding Agent, Smarter Assistance, Free Tier
The cost of inference[0] for the same quality has been dropping by nearly 10x year over year. I’m not sure when that trend will slow down, but there’s still been a lot of low-hanging fruit around algorithmic efficiency.

[0] https://www.reddit.com/r/LocalLLaMA/comments/1gpr2p4/llms_co...

drakenot··on Understanding R1-Zero-Like Training: A Critical Perspective
I've seen the same "Superficial Self-Reflection" mentioned in their linked blog post[0] as well, where the conclusion doesn't naturally follow the output of the thinking tokens. I think people are fooled by this, but if you take the time to inspect the "chain of thought" tokens they often don't match the final output answer.

I don't deny that performance for certain logic tasks goes up with these models but I don't fully understand what role the thinking tokens take in these cases.

[0] https://oatllm.notion.site/oat-zero

drakenot··on ZeroBench: An Impossible* Visual Benchmark for Contemporary Multimodal Models
Because of benchmark leakage / contamination?
drakenot··on DeepSeek not as disruptive as claimed, firm has 50k GPUs and spent $1.6B
Yes. And the original announcement and paper called out that this ~6 million was just for training the R1 model. It specifically mentioned that this didn't include any other costs, including research, previous models, etc. Just raw training costs for the latest model.

There is Hugging Face's "Open-R1" project which is attempting to replicate the DeepSeek-R1 model given the innovations outlined in the paper(s). They need to replicate the training code, the datasets, etc. This part of the system wasn't published.

And you are right that with the open weights of the R1 model, it should be pretty easy to infer how much it costs to _run_ as well.

drakenot··on Andrew Ng on DeepSeek
This just seems like Dario asking for regulatory capture and protection of his company. I don't think you need a deeper explanation than that.
drakenot··on Ask HN: Is AI converging on human-like cognition?
People have proposed doing CoT not with output tokens, but to do it in the latent space of the model.
drakenot··on Recent results show that LLMs struggle with compositional tasks
With DeepSeek-R1-Zero, their usage of RL didn't have reward functions really that indicated progress towards the goal afaik.

It was "correct structure, wrong answer", "correct answer", "wrong answer". This was for Math & Coding, where they could verify answers deterministically.

drakenot··on Open-R1: an open reproduction of DeepSeek-R1
What are some other domains outside of Math and Coding that would be suitable for RL with automated verification?
drakenot··on Why OpenAI's $157B valuation misreads AI's future (Oct 2024)
(Summary from Reddit)

- fp8 instead of fp32 precision training = 75% less memory

- multi-token prediction to vastly speed up token output

- Mixture of Experts (MoE) so that inference only uses parts of the model not the - entire model (~37B active at a time, not the entire 671B), increases efficiency

- PTX (basically low-level assembly code) hacking in old Nvidia GPUs to pump out as much performance from their old H800 GPUs as possible

Then, the big innovation of R1 and R1-Zero was finding a way to utilize reinforcement learning within their LLM training.

drakenot··on Nyxt: The Hacker's Browser
Every time I've tried to use Qutebrowser, or similar -- I end up bouncing from it because I want proper extension support.
drakenot··on Noi: an AI-enhanced, customizable browser
Are there any LLM products yet that can navigate websites for you ?

Or perform operations on your local computer?

drakenot··on Has Llama-3 just killed proprietary AI models?
Commoditize your complement and all that.
drakenot··on Show HN: Kyoo – Self-hosted media browser (Jellyfin/Plex alternative)
Bookcamp has fairly negative reviews on the App Store.

Have you had issues with it?

drakenot··on Ask HN: Where is Apple? They seem to be left out of the AI race?
Local Siri with function calling for dozens of on-device use cases around contacts, phone settings, interpreting web results, timers, calculator, calendar and phone manipulation.
drakenot··on Microsoft strikes deal with Mistral in push beyond OpenAI
The whole embrace, extend.. ?

> Mistral Remove "Committing to open models" from their website

That was 5 hours ago.

Without having insider details it is hard to know why, but the coincidence of timing with the Microsoft deal is not lost on me. It could have even been a stipulation.

drakenot··on The US military is embedded in the gaming world. Its target: teen recruits
Is there real evidence of people becoming more fragile?

Tons of young people still play sports and are physically active. Maybe it is just the bubble I live in, but kids who are on even rec leagues from 5 years old on are engaged in organized sport activity multiple nights a week.

drakenot··on Over 50% of the population of Germany does not own their home
Like the cabinets and whatnot? Wouldn't they be fitted to the space available or are the kitchen spaces standardized?
drakenot··on I designed a cube that balances itself on a corner
How effective would it be to build a neural network that controls the reaction wheel motors?
drakenot··on Show HN: Multi-monitor KVM using just a USB switch
Do you know of any solid KVMs with EDID emulation and 1.4 DP?
drakenot··on Neal Stephenson was prescient about our AI age
Have you been paying attention to the proliferation of AI / LLM generated content on the web?

Things are about to get dark unless we have new methods for combating it. It has seemed like Google has been on a slow and steady decline, losing this battle with spammers. This takes it to a whole new level.

drakenot··on Neal Stephenson was prescient about our AI age
Neal Stephenson's Anathem (2008):

> “Early in the Reticulum-thousands of years ago-it became almost useless because it was cluttered with faulty, obsolete, or downright misleading information,” Sammann said.

> “Crap, you once called it,” I reminded him.

> “Yes-a technical term. So crap filtering became important. Businesses were built around it. Some of those businesses came up with a clever plan to make more money: they poisoned the well. They began to put crap on the Reticulum deliberately, forcing people to use their products to filter that crap back out. They created syndevs whose sole purpose was to spew crap into the Reticulum. But it had to be good crap.”

> “What is good crap?” Arsibalt asked in a politely incredulous tone.

> “Well, bad crap would be an unformatted document consisting of random letters. Good crap would be a beautifully typeset, well-written document that contained a hundred correct, verifiable sentences and one that was subtly false. It’s a lot harder to generate good crap. At first they had to hire humans to churn it out. They mostly did it by taking legitimate documents and inserting errors-swapping one name for another, say. But it didn’t really take off until the military got interested.”

> “As a tactic for planting misinformation in the enemy’s reticules, you mean,” Osa said. “This I know about. You are referring to the Artificial Inanity programs of the mid-First Millennium A.R.”

> “Exactly!” Sammann said. “Artificial Inanity systems of enormous sophistication and power were built for exactly the purpose Fraa Osa has mentioned. In no time at all, the praxis leaked to the commercial sector and spread to the Rampant Orphan Botnet Ecologies. Never mind. The point is that there was a sort of Dark Age on the Reticulum that lasted until my Ita forerunners were able to bring matters in hand.”

drakenot··on jQuery v4.0 Beta
I remember when jQuery was first released. To convince my coworkers to use it, I held an internal presentation called "Don't Fear the $," where I walked through the advantages it provided.
drakenot··on 'I should not have written 'A Clockwork Orange''
I just remember reading The Stars My Destination from Bester and noticing all these elements:

* Mega-corporations with immense power and influence over all of society (Kodak of Kodak)

* Drug dens offering synthetic designer drugs that made people feel as if they were various animals.

* The protagonist receiving an upgrade to his nervous system, including a tooth-based switch that allowed him to 'speed up' for brief periods.

* Status symbols of the ultra-rich, such as using the most elaborate physical conveyances in an era of effortless teleportation.

I'm probably missing some, but it felt like such a prototype novel to something like Neuromancer and was written in 1956!

drakenot··on 'I should not have written 'A Clockwork Orange''
William Gibson is the father of Cyberpunk, not Steam Punk.

Though if Gibson is the father, I might say Bester is the grandfather.

drakenot··on Oasis – a small, statically-linked Linux system
You've just described Qubes OS!
← PreviousPage 2 of 11Next →