Hacker Newsnew | past | comments | ask | show | jobs | submit | natch's commentslogin

White nationalist is a wild, wild misreading though. Though not an uncommon one.

The Overton window doesn’t allow talking about all real issues, so people like yourself mistake race as the crux of the matter, when other issues are the real topic.

If you want understanding, work on helping open, not close, the Overton window.


Nah, he's shown his hand thoroughly at this point.

You may not realize this, but you're making excuses for his behavior. I'd recommend not doing that.


Very well reasoned and it fits a long evolved pattern but what you’re describing fits the mass sociopathic behavior called shunning, as engaged in by some religions like JW if you need a neutral example to think about. It is in no way good for society or for individuals in the long term.


Actual Nazis, sure. But don't believe everything you think.

Yeah, actual Nazis who commit administrative murder of a non-zero number of people in Africa who died from preventable diseases due to the DOGE USAID cuts. Some might call it administrative eugenics. I heard the guy who did that also threw a pretty obvious gesture that shows support for that ideology. I choose to believe my eyes and my functioning brain, thanks.

You bought a hardware device and you own that device?

This decoy notes scheme seems pretty unethical.


I wouldn't take claude research at face value. It found something odd, came up with a plausible sounding explanation and confidently presented it to OP. Doesn't mean it is true.

Sure, that goes without saying. That aside, if it is true, the point stands. The word “seems” already covered your point, fwiw.

Apart from model performance, what harness are people using to come close to Claude Code or Codex workflow styles with tool use, conversations, loops, remote control, etc.?


and more specifically - what harness is known to be the best fit for Qwen local models, and are there any evals/benchmarks for harness+model pairs?


I’ve noticed that when a developer cries foul over an Apple App Store rejection, it often seems to be a red flag that there is something fishy about that developer. Gruber, not to his credit, often does not give Apple the benefit of the doubt on these matters. imho he bears some responsibility if he feels duped.


...and I've noticed that's not always the case. What's the value of such a bland whitewashing comment?


Sure, glad to clarify. The value is to point out something that I’ve noticed a lot of people miss as conversations about App Store rejections tend to take a knee-jerk negative attitude toward Apple. Considering the opposite possibility would clear up a lot of false conceptions and raise the quality of discussions.


Nice work! I got lost in a clouds only view after setting my location but I’m sure I’ll figure it out, the first couple flights were over an unfamiliar area but worked fine.


Are you saying nobody should ever do anything with any tool if they cannot reasonably do it without the tool?

I mean personal blogs I agree the writing portion should be personal but does that go for everything? Maybe the author should take the emoji off his page(??!!)


He seems to be saying the cloud provider revenue has something wrong with it because it stems from serving training and inference. It doesn’t follow.


Cloud procider revenue is generally understood to be highly durable and safe because it is associated with both diversified customers and workloads which are expected to be reasonably constant in aggregate.

Zitron's point is thay OpenAI and Anthropic have generated considerable sector concentration in GCP and Azure. If the LLM sector were to contract in volume, this concentration could harm cloud providers financials more than what would normally be expected based upon their prior revenue streams.


27% of GCS revenue this year was from OpenAI and antropic, last years was 47%. With all the cross self dealing, this isnt a revenue source thats going to stay stable.


Title should not use a minus sign before the dollar amount, that’s super confusing.


Corrected thanks!


Sorry for the meta comment but: Why is such a useful comment “flagged dead” on what appears to be an interesting and pertinent post?

dang, any comments on what could be going on here?

I get it that there may be some indirect promotion happening but that’s natural when information is shared and this doesn’t seem like egregious spam.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: