HNHacker News
TopNewBestAskShowJobs

alchemist1e9

1,365 karma · joined April 30, 2022

submissionscomments
alchemist1e9··on GPT-5.6 Sol Pricing Cut by 50%
I agree. It’s flagged me on discussing fast GEMM implementations for large regressions, a discussion on theoretical physics math, a discussion on designing a type of RAG system. I’m super baffled as to what the safety instructions actually are other than “advanced anything” and even then their definition of advanced is a joke, I’m an idiot and my questions are almost laughable.
alchemist1e9··on GPT-5.6 Sol Pricing Cut by 50%
Fable is almost unusable for anything but super boring mainstream stuff. I was getting safeguard flagged so often I’ve significantly reduced my usage out of fear they will blacklist/ban me.

Some of the topics it’s flagged have been hard for me to understand what it seeing that can be remotely concerning in my requests.

alchemist1e9··on How art invented humanity
Fits the archaeological gap also because Homo erectus controlled fire yet left no art, clothing, or deliberate burials.
alchemist1e9··on Software for One
What?
alchemist1e9··on Mechanistic interpretability researchers applying causality theory to LLMs
Here is Yuji Tachikawa from Japan (Mathematical Physics, String Theory, QFT) on recent progress in his own work using Fable 5 :

"I've been trying out Claude Fable recently, and last night, on a whim, I showed it my research notes about a collaborative project that's seen no progress in the past six months or so and asked for its thoughts. To my surprise, it made a non-trivial observation and essentially solved it."

"I was also surprised that it was using sympy to automatically write code and verify his own predictions."

"Fable probably seems like it properly understands string theory and has intuition too—that's my impression"

alchemist1e9··on Mechanistic interpretability researchers applying causality theory to LLMs
It’s curious how they solve unsolved math problems without reasoning. Maybe I have a different definition of reasoning than you.
alchemist1e9··on Grok 4.5
I’d be curious to hear more about your dev setup and what tips you have for other aspiring vibe app coders.
alchemist1e9··on Micro-Agent: Beat Frontier Models with Collaboration Inside Model API
This should help with better utilizing a heterogenous collection of inference hardware.
alchemist1e9··on Pledging another $400k to the Zig software foundation
And how is it driving human extinction? It’s lifted humans out of endless poverty and bringing a coming age of abundance, what world are you living in?
alchemist1e9··on Pledging another $400k to the Zig software foundation
Capitalism is the single greatest force for improvement in the human condition. Money is the foundation of capitalism.

If you went to school and believe what you write then you went to terrible schools.

alchemist1e9··on Pledging Another $400k to the Zig Software Foundation
what was your original comment? I’m pretty sure it was a lot more critical sounding.
alchemist1e9··on Pledging another $400k to the Zig software foundation
Nah this is definitely bad type of envy you can look at the full discussion and additional comments from the user.
alchemist1e9··on Pledging another $400k to the Zig software foundation
EDIT: comment was under incorrect parent. my error. moved it to correct location.

EDIT2: Actually it’s more interesting. The commenters seem have changed their wording away from what I was criticizing.

Original observation: Try to purge envy from your heart. It’s a poison.

There was originally a lot of dark envy in this thread but interestingly it’s been revised out to be more subtle.

alchemist1e9··on GLM 5.2 Is Out
Could be but then it means like 98% is pornography I guess, because it’s every row, so if random a bad sign!
alchemist1e9··on GLM 5.2 Is Out
Why is the text field in dataset preview table populated with pornographic labels?
alchemist1e9··on Show HN: FablePool – pool money behind a prompt, and Fable builds it in public
Cypherpunks will be proud once there is a version of this cryptocurrency funded to providers receiving the cryptocurrency.

Or maybe there is? or a version where only those funding have access to the results.

alchemist1e9··on Ahoy, DECmate II the little PDP-8 that could
I was also thinking about the prices and what problems they were being used for to motivate the investment.

It then occurred to me that loaded Mac Studios and DGX Stations have some comparability in CAPEX scale. Here are some other prices for example:

> The VT278 started at $6,795 [$23,700].”

> This was sold as the DECmate III+ for $5145 [$15,400] alongside the standard III.

> The VAXmate finally hit the market in September 1986 starting at $4045 [$12,100].

> For the back end DEC announced a turn-key MicroVAX II system with 5MB of RAM, Ethernet, 16 ports and a 30-seat ALL-IN-1 plus WPS wordprocessing starting at $81,160 [$243,000].

alchemist1e9··on I turned a $80 RK3562 Android tablet into a Debian Linux workstation
> PostmarketOS is amazing on supported arm chromebooks.

Any tips on best models that are abundantly available used on the cheap and work well?

> I have a few that I throw in a bag for beach/jungle holidays - they are literal e-waste, something liberating about carrying a laptop that's worth significantly less than a decent family meal.

I definitely do this with a few Thinkpad 11e I have laying around from a failed project 4 years ago.

However I’d really like to switch to e-waste as what you describe would be very liberating. An e-waste Linux device with encrypted disk that you just wifi tether to phone and works fine for use old school types. I wonder how cheap they can go? How easy to flash? etc

alchemist1e9··on I turned a $80 RK3562 Android tablet into a Debian Linux workstation
It’s a great example and I have recently been thinking a lot that AI assistance maybe enable rapid porting progress and bringing life to recycled devices for 3rd world situations.

Linux can be trimmed way down and with an efficient stack on top can make many devices extremely useable.

Here is a related comment on user software side I made recently.

https://news.ycombinator.com/threads?id=alchemist1e9#4800737...

alchemist1e9··on Principles for agent-native CLIs
Well we will have to agree to disagree because my understanding of what has been generally the case is that the LLMs might vibe-coding spam, that’s true, but the interesting difference is generally speaking their “suggestions” are very reasonable and represent in hindsight useful changes that make the commands more useful for everyone, humans included.
alchemist1e9··on Principles for agent-native CLIs
I don’t remember exactly the specific examples off the top of my head (some are definitely ffmpeg commands) but I do know that when LLMs keep hallucinating command line flags that don’t exist for that specific command their “suggestion” is actually very reasonable and so many developers are adding support to their tools for common hallucinations.
alchemist1e9··on Principles for agent-native CLIs
It’s also likely that agents would also be better if they didn’t deal with json vomit either. I’m optimistic that agent frameworks will eventually come full circle and realize concise teletype linear CLIs aka old school UNIX is actually very effective and efficient for agents as well as humans!
alchemist1e9··on NPR finds "no sign" of Polymarket at its Panama HQ address
Does anyone have a good source that details these negative effects? I’m not doubting they exist, I mean gambling in general has many negative externalities, but I’m just interested in identifying the cancer aspects more specifically.
alchemist1e9··on GameStop makes $55.5B takeover offer for eBay
You mean leverage/borrowing? Pretty time tested mechanism of risk taking in free markets.
alchemist1e9··on A desktop made for one
You’ve reduce the memory requirements so much that it could all run on an early 90s computer easily. When I see such extreme examples I think back to the OLPC machines and this idea of how can extremely cheap but with useful software computers be available in very impoverished areas. I understand this has nothing to do with your argument or anything you’re writing about. It just made me think if LLM assisted software production might make the failed OLPC idea viable again. Could a minimalist but useful set of tools be created to run on old chromebooks for example.
alchemist1e9··on GameStop makes $55.5B takeover offer for eBay
> garbage capitalism.

How is this defined?

alchemist1e9··on A couple million lines of Haskell: Production engineering at Mercury
> I’m convinced the only reason people don’t use Mercury is that they don’t know what they’re missing.

Very well could be true because I had no idea who or what they are.

Do they have strong low level automation support for the customer programmatically even for personal accounts? I use ledger for plaintext accounting for both personal and business and sync of data is slightly annoying, perhaps Mercury’s products solve that trivially?

alchemist1e9··on High Performance Git
They wouldn’t be able to publish this useful knowledge easily without it though. And it’s the author’s guidance and vision which the LLM just helps materialize and so I think we should be studying how to generate content with less “slop” features and make it more natural and satisfactory for human readers, not discouraging it.
alchemist1e9··on High Performance Git
I think it’s great and you should be doing it, I have no problem at all if there is LLM assistance in authoring, I think it’s a good thing because like you said it enables solo writers with good ideas to produce valuable work that they otherwise wouldn’t!

What I’m interested in is how to address the “grating” or whatever characteristics the readers detect to have them focus on the LLM aspect. I feel it’s probably soon or already removable with some methods.

Ignore the haters they are just wrong to blanket criticize, however their observations are helpful to try and improve the process. We want LLMs to assist in creating useful and effective content for humans.

alchemist1e9··on High Performance Git
I think it’s likely there will be methods to fix this soon, some de-slop algorithms, or is there a deep reason it will always be detectable? Perhaps there are some PhD linguists who have figured out how to quantify the “slop” effect and are writing their thesis on it. Once that is done it will be possible to smooth it away.

The book is definitely LLM assisted authoring yet it also has great content, so not sure we can immediately jump to shaming it entirely for being slop.

← PreviousPage 2 of 34Next →