HN user

randomperson_24

352 karma
Posts52
Comments62
View on HN
research.google 9mo ago

Speech-to-Retrieval (S2R): A new approach to voice search

randomperson_24
4pts0
www.theverge.com 3y ago

Spain bans setting the AC below 27 degrees Celsius

randomperson_24
58pts103
www.youtube.com 3y ago

Jigzilla: The automatic puzzle solving machine (part 1)

randomperson_24
6pts0
www.vimarsh.info 4y ago

Running a public ad-blocking DNS Resolver

randomperson_24
2pts0
www.wsj.com 4y ago

China Tops Google, YT Results on Covid Origins and Beijing’s Human-Rights Record

randomperson_24
5pts0
www.vimarsh.info 4y ago

Information, Bandwidth, and Trust

randomperson_24
1pts0
cloud.google.com 4y ago

Google Cloud unveils world’s largest publicly available ML hub with Cloud TPU v4

randomperson_24
1pts0
www.deepmind.com 4y ago

Dynamic Language Understanding

randomperson_24
1pts0
80000hours.org 4y ago

Climate change – Many think it's world's top problem. Is this right?

randomperson_24
9pts3
storage.googleapis.com 4y ago

A Generalist Agent [pdf]

randomperson_24
72pts8
tokcast-bretthall-31ddc0.netlify.app 4y ago

Are we running out of resources?

randomperson_24
51pts201
sec.okta.com 4y ago

Official Okta Statement on Lapsus$ Claims

randomperson_24
2pts0
deepmind.com 4y ago

Predicting the Past with Ithaca

randomperson_24
11pts0
www.vimarsh.info 4y ago

Running your own email server

randomperson_24
7pts0
www.0xvimarsh.com 4y ago

I hosted my website on a Raspberry Pi

randomperson_24
2pts0
fee.ninja 4y ago

Know how much you have spent on gas fees

randomperson_24
2pts0
www.theverge.com 4y ago

Intel apologizes to China for following US sanctions targeting Xinjiang genocide

randomperson_24
8pts0
www.0xvimarsh.com 4y ago

Things I found interesting last week

randomperson_24
2pts0
www.0xvimarsh.com 4y ago

Interesting things I found last week

randomperson_24
1pts0
www.0xvimarsh.com 4y ago

Interesting Things This Week

randomperson_24
1pts0
www.youtube.com 4y ago

YouTube Is Changing – Smarter Every Day 266

randomperson_24
2pts0
www.vimarsh.info 4y ago

It isn't Facebook, it's us

randomperson_24
1pts0
www.vimarsh.info 4y ago

Self host Apps using GUI – Caprover

randomperson_24
3pts0
aryantiwari.com 4y ago

To 1980, from 2021

randomperson_24
1pts0
beyondbubble.org 4y ago

Letter to 1980

randomperson_24
1pts1
www.theverge.com 4y ago

Waymo is bringing its autonomous vehicles to New York City for mapping

randomperson_24
1pts0
www.vimarsh.info 4y ago

Simple way to self-host apps

randomperson_24
1pts0
www.youtube.com 4y ago

Google May Cut Me Off on November 1 – Walter Lewin

randomperson_24
6pts1
www.aiweirdness.com 4y ago

Stealing a giraffe from the zoo? Only if it's a cool giraffe

randomperson_24
2pts0
www.vimarsh.info 4y ago

Easily self-host and manage multiple apps via a simple web interface

randomperson_24
3pts0

India doesn't have enough purchasing power to convince big tech companies (Google, Facebook, etc.) to store the data of Indian residents in India. Also, there are not many local tech monoliths with consumer apps to fall back on (like China had).

Since India depends heavily on apps & services by big tech, I doubt such requirements like GDPR has would be enforced. Although the right to ask a website to delete all information about one and so on are quite good and could work.

Apple Maps is honestly no where close to Google Maps. Especially in dense Indian cities with multiple roads to reach a place, some roads are not even there on Apple Maps. It also (like OSM) lacks in contact details of businesses, open/close time, etc. which are a real value add for Google Maps.

Hopefully some real competitor emerges.

I think Google can probably not fix it. Users will have to be manually reporting as spam. These websites on seeing traffic from Google's crawler bots show a perfectly legit and highly SEO optimized website, but for anything else show other spam. If Google starts indexing from random IP ranges, most websites would probably block indexing from “unofficial IPs” or some companies (esp in EU) would file some lawsuit against Google. The reason being that some pay-walled news article websites won't be indexed properly, as the “unofficial IP-ed” Googlebot will not get the paywalled content.

If a website lies to Google itself, I believe the only way to solve it is by reporting the search result as spam or Google contracts people to somehow visit all billions of web pages (again the same problem – from different IP ranges) to verify it as a legit page.

I would like to know how Google currently handles it and probably how it could be improved

I have tried playing around with the beta access to make it generate NFT art with different prompts, but in vail.

I think it has not been trained on NFT art (crypto punks and so on).

Then why did China ask for the Hambantota port[1]. It is a clear strategic and economic policy. Not to deny other countries haven't done that in the past; but it is a clear way of gaining control - especially the taking control over infrastructure part.

Same is the case with Africa, and many projects of which do not benefit the citizens of those countries in any way too. [2]

[1] https://www.nytimes.com/2018/06/25/world/asia/china-sri-lank... [2]: https://www.youtube.com/watch?v=aJSD8XV3qzE

I think the problem is that we all expect to get perfect results. We want automated systems to identify and remove unimportant useless things. But how will those systems get to know that?

We can't build another AI black box to try and do that. There is a option in search to report spam or improve search results.

Out of our laziness and expecting getting everything perfect, sometimes collectively working and reporting bad content-copying websites, etc. would be collectively useful for all.

Search had gotten so good that we probably expected to get perfect results everytime. Untill bad guys started to game it. I beleive it's more of cat and mouse game, and where all users collectively should help and mark as spam.

I think the problem is that social media platforms, especially Twitter downranks your post if it has a link.

Therefore, recently people have been writing threads with a link to post at end of thread. The reason simply is that platforms want people to remain on the platform for longer periods of time.

Instagram doesn't even allow you to have a link in your post. Just in the profile.

Unpopular opinion: Android is more privacy focussed than iOS.

Especially because during first setup, it clearly asks you to opt-out of services you do not wish to participate in. Also much more transperent.

This is highly inaccurate. The author is using Plausible via a proxy but Google Analytics directly (a very biased way to do things).

Secondly, PiHole, uBlock origin and most other adblockers also block Plausible analytics (there are discussions on their own Github Issues regarding it). If you are proxying Plausible to get accurate readings, same should be done with GA.