Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
foobiekr
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
121.
▲
by
foobiekr
9mo ago
Wine is more stable as an API to target than any of the native Linux technologies.
122.
▲
by
foobiekr
9mo ago
The problem is that software people are clueless.
123.
▲
by
foobiekr
9mo ago
And if France decides to jam you until you play nice?
124.
▲
by
foobiekr
9mo ago
Space based power arrays with microwave transmission to massive ground fields has been discussed for nearly sixty years. It doesn't make economic sense, at all. https://caseyhandmer.wordpress.com/2019/08/20&#x
125.
▲
by
foobiekr
9mo ago
The lifetime of a GPU in practice is around 3 years. There is no way this plan is going to work. Musk knows that, he's just counting on stupid people to buy into the SpaceX IPO on hype.
126.
▲
by
foobiekr
9mo ago
Musk lies with ease and routinely. This and Optimus are both just more examples of that.
127.
▲
by
foobiekr
9mo ago
A lot of acquisitions are tech and talent and the details matter. A startup that gets acquired for $30M after burning $70M is definitely burned.
128.
▲
by
foobiekr
9mo ago
They aren't just showing people how to use their software. Stop defending the bad guys.
129.
▲
by
foobiekr
9mo ago
Palantir does a ton of customization and consulting for specific use cases. This isn't like Microsoft Excel being used to track uranium enrichment in Iran, it is a direct, explicit part of their business. Even if you do nothing else of
130.
▲
by
foobiekr
9mo ago
Due to background, I know a lot of people who work at google, and while many of them will give lipservice to ethical concerns, none of them have made any changes at all because, and this is an exact quote, "the money is too good."
131.
▲
by
foobiekr
9mo ago
This take is so wrong it qualifies as delusional. The valley was all about money and nothing but money by any means by no later 1996 when the dotcom got under way. In 2001 I was at a company actively engaging in meetings with a certain thre
132.
▲
by
foobiekr
9mo ago
Same. I would never allow anyone who has Palantir on their resume to be hired in any company I have influence over. They are the brownshirts of the tech industry, worse even than the people poisoning children's minds at Meta.
133.
▲
by
foobiekr
9mo ago
Once it’s in the context window the model invariably steers crazy. Llms cannot handle the “don’t think of an elephant” requirement.
134.
▲
by
foobiekr
10mo ago
The NTP server thing was especially egregious given that the transition to everything being under TLS was underway and clocks matter in that situation.
135.
▲
by
foobiekr
10mo ago
"What happens when multiple devices in your /8 want to listen on port 80 and 443 on the public address?" This is a feature not a flaw. The average person doesn't have anything acting as a server, and that's a good t
136.
▲
by
foobiekr
10mo ago
> would never be able to perform well on novel coding or mathematics tasks. We were wrong I'm not clear at all we were wrong. A lot of the mathematics announcements have been rolled back and "novel coding" is exactly where
137.
▲
by
foobiekr
10mo ago
I am sorry that has been your experience. I have worked in a lot of "rough/gruff/hardcore" environments, almost all of my career, at companies that are widely recognized to be fairly political and antagonistic, and none
138.
▲
by
foobiekr
10mo ago
What other option do they have? This is a "the horse might sing" situation for the whole market that focused on breakthrough-level results (AGI, ASI, or even just "not going off the rails after the third response").
139.
▲
by
foobiekr
10mo ago
Agreed. A good example might be one that came up recently (on "Tasting History") is musk, which tastes of and has the bouquet of laundry detergent, or basil (which if you eat it plain and in isolation has a bizarre chemical flavor
140.
▲
by
foobiekr
10mo ago
I have done the same; it's not natural to do it this way. Go should actually express an explicit mechanism to do this. When I did it, it felt exactly like trying to use epoll from Go: you can do it, it just feels like crap.
141.
▲
by
foobiekr
10mo ago
It's an obvious cost optimization. Make the consumer directly cover the cost of inference and idle inference hardware.
142.
▲
by
foobiekr
10mo ago
Two big issues in Golang are that you can't actually build an arena allocator that can be used for multiple types in a natural way. The other is that almost no library is written in such a way that buffer re-use is possible (looking at
143.
▲
by
foobiekr
11mo ago
It really wasn't. MacOS classic was full of vulnerabilities as was OS/2 and Linux up through 2004. Windows dominated because it was the biggest ecosystem.
144.
▲
by
foobiekr
11mo ago
"Type 3 diabetes" is one of the speculated causes of alzheimer's. The evidence there is not great.
145.
▲
by
foobiekr
11mo ago
Even the people on comp.arch at the time were baffled. No one believed it.
146.
▲
by
foobiekr
11mo ago
Amazon had paper losses and great cashflow. OpenAI has actual, real losses and terrible cashflow.
147.
▲
by
foobiekr
11mo ago
And the government backstops. Nothing says confidence that AGI is imminent like needing the US government to prevent your investments from losing you money.
148.
▲
by
foobiekr
1y ago
Umbrella seems to be blocking it, for one.
149.
▲
by
foobiekr
1y ago
"If you described" Yes, and if they used it for awhile, they'd realize it is neither general nor intelligent. On paper sounds great though.
150.
▲
by
foobiekr
1y ago
You'd have to have some incredible rapid attrition to have the median be above the average. Any tech company that left skewed would be considered a disaster by pretty much anyone.
More ›