Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dartharva
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
91.
▲
by
dartharva
4mo ago
On a side note, am I the only one who feels Codex models have a higher general first-pass success rate than Claude models on coding what you want? I use Github Copilot and always find myself drifting more towards them when working.
92.
▲
by
dartharva
4mo ago
Indeed, for work and software most are already beholden to Microsoft and Google. This is something wayy more.
93.
▲
by
dartharva
4mo ago
Was about to ask the same
94.
▲
by
dartharva
4mo ago
Could just be a courtesy - Americans tend to be rather suspicious and hostile to contributions coming from China, and it might draw unwarranted attention from agencies and bad media.
95.
▲
by
dartharva
4mo ago
I reckon such AIs are already in place, but by proxy. There must be CEOs somewhere who have completely offloaded what meagre amount of thinking they needed to do to some bespoke AI setup while they LARP around convincing people they are the
96.
▲
by
dartharva
4mo ago
Generally, yes. Also the gap between the employee and executive class is a lot smaller instead of unnaturally inflated like it is in most private equity companies.
97.
▲
by
dartharva
4mo ago
Coops can definitely succeed and even dominate at large scales with some minimum government protections. The largest dairy producer in the world is a coop: https://en.wikipedia.org/wiki/Amul
98.
▲
by
dartharva
4mo ago
I wonder if business schools could ever start actually teaching this skill. So far they just largely operate as scams, held up purely by prestige and network value
99.
▲
by
dartharva
4mo ago
Jeez, what a hostile crowd. I personally am just glad companies like Microsoft, that infamously used to charge for every material they put out, have been providing free learning resources to newcomers like this, quality notwithstanding. I a
100.
▲
by
dartharva
4mo ago
wth are you talking about Isn't the mere fact that every HN frontpage is filled with AI-related articles not indicative enough of how much it holds interest here? > post about how AI “writes bad code,” “introduces bugs,” “creates te
101.
▲
by
dartharva
5mo ago
India never had that choice, because forced order in the magnitude of China is practically impossible for the country for a vast variety of reasons too many to list down. Suffice to say there can be no comparison between PRC's federal
102.
▲
by
dartharva
5mo ago
There is no singular decisive answer. Sure, you can chalk all of it up to "wrong" policy, but peel the layers a bit and explore the causes behind them and you'll see an endless chain of cause-and-effect covering more factors
103.
▲
by
dartharva
5mo ago
Even if they are, it still doesn't justify the ridiculous levels of overvaluation. They are not essentials and their consumer demand is extremely elastic.
104.
▲
by
dartharva
5mo ago
Had this assertion been true India wouldn't have been the most populous country on earth.
105.
▲
by
dartharva
5mo ago
Please.. undergrad college in any stream is a very achievable baseline that literally anyone not afflicted with a pathological mental condition can pass, provided they are interested themselves and are subjected to classes from instructors
106.
▲
by
dartharva
5mo ago
More than the CEOs themselves, I feel news media are more suffering from "AI psychosis" than anyone else with the constant stream of FUD articles they keep churning out.
107.
▲
by
dartharva
5mo ago
Just asserts "Tech CEOs" in general, I don't see any names being named. Meaningless FUD article
108.
▲
by
dartharva
5mo ago
This↑ - I had also jumped ship to DDG until I realized this query string still works on Google.
109.
▲
by
dartharva
5mo ago
This article is a whole lot of nothing, feels like just an excuse to feature Goldman CEO on the publication for prestige value.
110.
▲
by
dartharva
5mo ago
Surprising that nobody's talking about the non-AI Google - https://tedium.co/2024/05/17/google-web-search-make-default/
111.
▲
by
dartharva
5mo ago
The captcha on this site is irritating. Original paper: https://www.bmj.com/content/332/7536/266
112.
▲
by
dartharva
5mo ago
> while Western Democracies are afraid of the future, and of their own shadow. Trillions of Dollars being invested against AI infra would indicate otherwise. US is in fact betting a lot of its economic future on AI.
113.
▲
by
dartharva
5mo ago
This is also the kind of advice that only sounds good on paper. In reality there is no clear marker of what you can and cannot do well unless you empirically experiment with everything which would take several lifetimes. Modern gurus like C
114.
▲
by
dartharva
5mo ago
DuckDuckGo gives you enough control in its settings to completely filter out AI-anything in search results. I like to think it is near classic Google's levels in search result quality even though it probably isn't.
115.
▲
by
dartharva
5mo ago
Distro wars are pointless when the main point of friction remains the upward fight that is getting useful software to run on Linux.
116.
▲
by
dartharva
5mo ago
You are not supposed to use Windows 11. Use any of Windows 10's enterprise versions, the difference is night-and-day.
117.
▲
by
dartharva
5mo ago
This completely resonates with my experience. I have amassed enough familiarity with Linux having used and tried out all kinds of distros to the point of being able to assist others in setting up and diagnosing their systems, but I myself k
118.
▲
by
dartharva
5mo ago
The video version of this article made me feel very good. Not just AI anxiety, it shares a template of better living in general. Highly appreciate.
119.
▲
by
dartharva
5mo ago
Will this get LTSC's security updates? I can seriously try this out as a daily driver if yes
120.
▲
by
dartharva
5mo ago
All he brings up are social media comments, which may have been mildly representative of people more than a decade ago, but in the present-day are widely understood to be completely immaterial.
More ›