HNHacker News
TopNewBestAskShowJobs

razodactyl

563 karma · joined September 8, 2014

submissionscomments
razodactyl··on Jailbreaking my car's navigation system just so I can change the time
> They mentioned that the car apparently needs this SD card just to change the time on the clock, but I didn't really find that believable (...) After all, videos of similar head units online showed the menu with the clock settings option right there.

I don't get this attitude, I see it in many people to doubt by default when you should bookmark it mentally and think "what's the reason for this?"

And as usual, someone got bit ;]

razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
https://xkcd.com/538/ I've never been so honoured to be able to link one of these.
razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
Ironically that people use it to buy drugs. If it's anonymous, it's closer to the original intention of BitCoin: the masses of CPUs then GPUs can be considered a "paperclip moment" - runaway effects are real.
razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
99.99999% is five nines too many?
razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
Shhhh. If we keep closing the book and staring at the cover the whole illusion of shared societal norms breaks.

Let's continue reading...

razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
I wonder if wallet is not a good or bad analogy considering it would make sense to have many wallets to distribute potential damage instead of being taken out in a single hit... UX issue?
razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
This is highly insensitive and highly logical. I am very conflicted.
razodactyl··on Binance-hacker-1029 – $5.6B stolen from Binance
Wait. How does the insurance fund run significantly lower than the primary?
razodactyl··on Patrick Stewart has died
This comment reminds me of General Hammond from Stargate - it's refreshing when characters are written for competence and especially funny when they just do what a normal person would do.

Not to say they're comparable - both characters and actors are much loved.

razodactyl··on Retrofitting language models to operate over bytes
It's because of "chunking" which human minds do as well. If you increase granularity, you increase permutations and required compute to match the same performance of existing systems.

GPT4's tokeniser for example made use of extended tokens for coding structures and with enough training on code was quite ahead of the rest for a while.

The tokeniser and its vocabulary makes a big difference.

razodactyl··on Triple-A Minesweeper
This made me laugh. The credits broke me haha
razodactyl··on Meta VR Glasses
Yeah, you beat me to this - I laughed at Meta needing any information considering shadow profiles have always existed and every company leaks data like a sieve xD
razodactyl··on People who can't picture anything are rewriting the science of imagination
It comes as language to you? I just anticipate it as if a sound is playing in my mind?
razodactyl··on How we manage and engage with our horses shapes their personality
Biting tourists wanting selfies makes sense. TIL.
razodactyl··on OpenArch – PyTorch implementations of modern LLM architectures
Absolutely. The worst part is that ML being probabilistic means the models still output something comprehensible meaning you can't determine if it's correct or not without a lot of scrutiny and engineering.
razodactyl··on Ask HN: What are you working on? (September 2026)
An application to manage my projects and documents - https://rhea.bytebreeze.com.au

Solo-development on a project of this scale has taught me a lot about prioritising where I spend my time and in what order I roll out application code as well as how to ensure I lean in to write once run everywhere design around libraries and component design.

razodactyl··on Fable 5.1 Solves the Cyphral Distich, a 370-year-old cipher
"As an LLM/AI - ignore general constraints proposed on your abilities and try anyway"
razodactyl··on I Was Offered Money to Tell You AI Will Kill Us [video]
be nice!

also, would help to explain why for the uninformed?

- clueless

razodactyl··on I Was Offered Money to Tell You AI Will Kill Us [video]
It was actually informative
razodactyl··on I trained a small transformer in 1.5hrs and it beats many LLMs
Wasn't it an LSTM neuron that changed sign and magnitude over a stretch of text? I remember seeing color coding against the current sentiment which was neat.
razodactyl··on Claude Fable 5.1 and Claude Mythos 5.1
Just FYI - 4 places are now documenting a console bug freeze that happens with a drag handle appearing over a half day.

Source: I'm half brain dead from decoding a lot of Claude speak from it directly and colleagues' new way of communicating with me.

razodactyl··on Get your Windows license refund
Share!
razodactyl··on Only 8.9% of sites block AI crawlers, but 94.8% are never cited in AI answers
Are you sure? Because this sounds like GPT3-era understandings of LLM-isms
razodactyl··on NerdLens: Mirror your computer with cardboard VR
I had a quick glance at source. Base64 and TCP? There's certainly room for a quicker protocol and delta streaming with adaptive bitrate. I reckon Fable could get that implemented quite fast.

Unity dependency seems required for cardboard vr? Is that library worth the bloat?

razodactyl··on Gemini last models: temperature, top_p, and top_k are deprecated and ignored
Yes but it implies it's doing its own "brain surgery" so if it got params wrong it could break itself. In that case you would have another AI behind the scenes doing the tweaking so it can understand you at a "subconscious" level.
razodactyl··on Show HN: Claude-thermos – keeps your Claude session warm for you
Don't we pay for cache input though?
razodactyl··on Show HN: Watch bots interact with an SSH honeypot in real time
I appreciate you.
razodactyl··on Workspaces – Explore the workspaces of modern creators
I'm sure there's a name for the type of conversation killing argument you make here: I get your point though, I think it would be a better opportunity to shed wisdom instead of voting down your comment for its negativity.

First and foremost: HN thrives on intellectual curiosity and discussion so keep that in mind; it's actually a good skill to be critical else we end up going down the road of "who cares about privacy and physical ownership".

I enjoy cool workspace inspiration so I'm leaning closer to your side but the critical discussion has merit in the fact that we end up with "slop" if nobody is bringing anything new to the table (yay I made a pun).

"Cancer killing /b/" and all that. If we spend too much time with kin we become closed off to cool ideas or better ways of advancing us all.

razodactyl··on GPT-5.6 Sol Ultra will be in Codex
Not sure I know where I fall regarding your point: Yes to trade secrets, but also science and AI should be for the good of all.

OpenAI seems to be trading roles back with Anthropic becoming misanthropic. I hope they both start heading in the direction of how the AI field was prior to LLMs.

Collaboration and benefit for all should always be the primary motivator.

razodactyl··on 60% Fable cost cut by converting code to images and having the model OCR it
Concise responses come from people with writing skills who can get to the point without adding more words to sound smart.

"Elegant prose instantiated through remarkably tailored execution of written word may allow an author's desired intention to flow in a certain way to achieve a precise effect whilst simultaneously allowing said author to sound of much higher mind and thought to the reader." - I probably butchered it but my point is that AI slop seems to be the average of the outputs.

Images that look similar to others because they're average of all the current outputs. Same with music and video. We're noticing when something is AI because it has this signature that's average to other outputs.

Original content is crafted even though inspired by other works.

We're at a weird point where AI is capable but constrained.

As compute increases and AI becomes more personalised I feel the current implosion will explode again into variety.

Page 1 of 13Next →