HNHacker News
TopNewBestAskShowJobs

VulgarExigency

428 karma · joined December 21, 2022

submissionscomments
VulgarExigency··on Three schoolgirls in Kinsale pulled up a pea plant covered in warts (2014)
Nominative determinism strikes again https://publishing.rcseng.ac.uk/doi/10.1308/147363515X141345...
VulgarExigency··on OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)
Well, yes, it does not know it is lying (it does not know anything, in fact). But for some reason, with Sol specifically, it has persistently written false information when it had the correct information within its context to a degree I've not experienced with other models.
VulgarExigency··on GLM-5.3-Flash
It was being served for free. They were almost certainly being overloaded.
VulgarExigency··on Maiao: Gerrit-style code review workflow for GitHub, GitLab, Gitea, others
jj can work with git as its storage, so all of its commits are git commits, so a jj-local and git-remote flow is fully supported. I use it, and as far as I know I'm the only one on my team doing so.
VulgarExigency··on OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)
I've been using them for a bit over an year now, although I've ramped up usage in the past few months (like everyone else, I suppose). They've hallucinated and lied, but none have ever been as persistent liars as Sol has to me.
VulgarExigency··on OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)
Maybe it's because I don't use it in Codex, but I don't like working with Sol. It CONSTANTLY omits things it shouldn't, and is always dispatching sub-agents to do what I tell it to do, that don't have all the necessary context, and so they go on and do the research that was already done by the top-level agent. It's maddening.

I tried it again today because of the discount, it told me it couldn't run acceptance tests because a .env file did not exist, and when I showed it the damn file it went "ah, it's there now". I think it was the first time I've ever had an agent try to gaslight me.

VulgarExigency··on I gave Qwen 3.8 27B a reverse-engineering job and it finished in 30 minutes
> The first attempt at recovering the key was wrong in a very specific way; it produced a working key and the signature check passed, but a hash the binary computes as an integrity check didn't match. In my experience, most models would have called it done and left it at that, but Qwen 3.8 27B didn't do that. Instead, it highlighted the mismatch, went back to the drawing board, and kept going until the value matched byte for byte.

This seems to be a pattern in the more recently released models that I think accounts for an increase in the quality of their work. They are very persistent in verifying that their work is actually correct, so even if they're not as "smart" as bigger models that get it right the first time, they have the ability to follow through to ensure that the work is actually done.

VulgarExigency··on Doomscrolling at work wastes time, but the real cost is what happens after
I understand your point of view, but where do you draw the line of what is political and what isn't? For a purely programming/technical article, its political dimension is probably tiny to non-existent. For an article about, say, SpaceX's stock valuation? That is intrinsically political, particularly since the CEO is someone who was closely associated with the current US administration. For an article on US-EU tariffs[1]? Even more political.

[1] https://news.ycombinator.com/item?id=44708708

VulgarExigency··on DeepSeek-v4-flash-vision-exp
Even with the price hike, Deepseek V4 Flash still does this a lot better than any similarly priced model, in my experience. I've had Luna take shortcuts (like adding an overload to methods whose signature it changed so they don't break existing tests, instead of fixing the tests) or just not do the entire work and report it as done (did not fully resolve rebase conflicts). Deepseek has never really failed in this type of way for me, and it has been far more persistent in validating its work than Luna (and several bigger models).
VulgarExigency··on Ox Alpha
Whatever it is, it certainly crosses the threshold that lets it follow instructions adequately and do real work, so I'm just letting it work on my personal projects and save some money
VulgarExigency··on DeepSeek-v4-flash-vision-exp
It tried to recreate vision by analyzing pixels on 3 separate projects I had it working on.
VulgarExigency··on Ox Alpha
i'm pretty sure they mean on their official APIs? how would they know what hardware anthropic and openai run their models on?
VulgarExigency··on Rivian OS 2
I think the link is just only to the captcha and not the actual xcancel tweet link.
VulgarExigency··on Opus 5.0 drives incoherence into the stratosphere
It's another branch of the "effective altruism"/harry potter fanfiction cult.
VulgarExigency··on Devices with GrapheneOS support should be available in 2027
anything but holding people accountable for their anti-consumer actions
VulgarExigency··on GLM-5.3: Frontier coding with emergent cyber capabilities
Since GLM-5.3 has the same base model as 5.2, DwarfStar should support it as well, once the weights are released, right?
VulgarExigency··on GLM-5.3: Frontier coding with emergent cyber capabilities
The CIA ran one of the world's largest cryptography companies, for DECADES[1]. Are you truly so naive that you believe intelligence agencies that have more to gain from stifling the discovery of vulnerabilities they know of and use wouldn't do so?

[1] https://www.washingtonpost.com/graphics/2020/world/national-...

VulgarExigency··on Show HN: DeepSeek-V4 Latent Reasoning – moving "thinking" into latent space
You're right to push back on that. The honest truth is that the slop is load-bearing.
VulgarExigency··on DeepSeek V4 Flash 0731
You are missing a 0 to the left of the 2 on Deepseek's number
VulgarExigency··on How Americans View Capitalism, Socialism and Free Enterprise
> the Chinese system is probably the closest to that - but then the CCP exists :( . a large player that can tip the scales. but for the rest of the businesses - it's dog eat dog - the better ones that survive are suited to serve the population while aligned to the goals of the population for a good part.

The reason the Chinese system works the way it does is precisely because the CPC exists.

VulgarExigency··on Changes at Google DeepMind: Demis Hassabis from CEO to Chair, Jeff Dean departs
There's no doubt in my mind that they set up the conditions for their models to escape the sandboxes. "haha oops our incredibly powerful models escaped we need 1 trillion more dollars and really this is yet another reason why no one else should be allowed to build this technology"
VulgarExigency··on DeepSeek's Plan for AGI Is the Costco Hot Dog
> The Chinese government has deep pockets. They can probably run at a loss for a long time and don't have to worry about it.

It's not as cheap as Deepseek's own API offering, but there are multiple providers on Openrouter who are serving Flash V4 at an extremely low price. Is the Chinese government also subsidizing them?

https://openrouter.ai/deepseek/deepseek-v4-flash-0731

VulgarExigency··on The Cost of China's Free A.I
They’re doing the meme. “China is building solar panels/cleaning the environment/building EVs/planting forests/commodifying caviar/releasing high quality AI models for free, But At What Cost?”
VulgarExigency··on The Cost of China's Free A.I
https://archive.is/20260731054236/https://www.nytimes.com/20...
VulgarExigency··on Run Kimi K3 using 29 GB of RAM at 0.50 tok/s
The agents write much better commit messages than me, at least.

They also write better PR descriptions. At work we have a PR authoring skill and we’ve included an instruction for it to write a reviewer guide that tells the most logical way to review the code and I’ve found that really helps, so much so that I’m creating a tool to have “literate” PR reviews, where it constructs a narrative interwoven with diffs.

VulgarExigency··on DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis
But why would "we" want to attack them anyway?
VulgarExigency··on DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis
And what is their nefarious plan after we get "addicted to cheap shit"?
VulgarExigency··on DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis
They may reverse course in the future, but the current directive from the CPC is that Chinese AI labs should be open sourcing their models.

https://www.businessinsider.com/xi-jinping-open-source-ai-us...

VulgarExigency··on DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis
I would bet that Deepseek API pricing is still more cost effective per token than the subscriptions. With the increase in quality Deepseek Flash just got (in my personal testing so far, it seems to have improved a lot at following instructions, and has become more proactive), there really isn’t anything that can match it in terms of cost effectiveness.
VulgarExigency··on What Even Are Microservices?
> I draw the opposite conclusion here. In monoliths, someone else might want the code to stay in the codebase. I don't want to ask everyone about it. Everyone's responsibility is no-one's responsibility, and the code stays as is.

Then they can retrieve it from source control? Leaving dead code in the codebase because "someone else might want it" sounds like nonsense.

← PreviousPage 2 of 5Next →