Abstract: How could machines learn as efficiently as humans and animals? How could machines learn to reason and plan? How could machines learn representations of percepts and action plans at multiple levels of abstraction, enabling them to reason, predict, and plan at multiple time horizons? This position paper proposes an architecture and training paradigms with which to construct autonomous intelligent agents. It combines concepts such as configurable predictive world model, behavior driven through intrinsic motivation, and hierarchical joint embedding architectures trained with self-supervised learning.
"Newsbeat is the BBC's radio news programme broadcast on Radio 1, 1Xtra and Asian Network. Newsbeat is produced by BBC News but differs from the BBC's other news programmes in its remit to provide news tailored for a specifically younger audience."
You explained to me (and HN) recently that posts which are critical of HN itself are moderated less, not more, than others.
The same standard should apply to posts in which a greatly disadvantaged group are standing up to a vastly superior power, all else being equal (credibility, tactics, nature of grievance, etc.)
As with content concerning the events of June 4, 1989, in Tiananmen Square, China.
Dupes should be merged. Valid freestanding posts should be unflagged.
Ok, I've removed it. Now people will accuse us of rewriting history, but that's ok - one eventually makes peace with the fact that there's no way to win.
dang says elsethread, paraphrased, 'we evaluated that flagged post and figured it was probably legitimately flagged, but after users contacted us^ to say it should be unflagged, so we listened to them and unflagged it'.
Good intentions often lead to bad outcomes. The mods have the best possible (and still imperfect) judgement of all of us, and they misread the circumstances, same as those users who flagged it. So they'd have to ban themselves right along with everyone else, which would be the end of HN.
That the actions of other users ended up being reversed is no excuse to demand that they be banned, especially on topics where agreement will never be reached^^ by the community — such as Apple, China, or Firefox. Please be more cautious about demanding bans of those whose actions result in disagreeable outcomes, even if you doubt their intentions.
^ Using a post or comment is a terribly inefficient way to go about doing this, but it certainly does tend to produce maximum-drama results. I don't think anyone should be banned for taking that approach — maybe they didn't realize the Contact link exists or reaches the mods! — but I'm exhausted of otherwise good-standing and high-karma users using posts and comments to bring things to their attention. I'm also glad to see the thread and discussion here, which contradicts my own viewpoint. I have no easy resolution for that cognitive dissonance.
^^ Please don't call for me to be banned if you disagree with my reply. However, if you feel I've violated the guidelines in some way, I do of course encourage you to let the mods know using the footer Contact link.
EDIT: HN's contact email is working fine; the evidence presented was known to be invalid before it was posted and used in bad faith to bait the conversation.
It worked fine when I emailed them yesterday using a non-Gmail email client and server. You’re suggesting that all email for YCombinator is down (or, worse, some/all of Google Apps) so that seems worth proving out. I checked and there aren’t any posts today about Gmail or Gapps being down. Diagnostic superpowers, activate!
Are you getting a bounce/timeout from a real email client through a real SMTP server? I know some mail servers I used to operate would reject your telnet test as shown, so I’m not really surprised it failed at a Google SMTP endpoint.
Are you attempting to email them using Gmail, using a third-party commercial service, via a cloud/datacenter-hosted personal SMTP server, or a home internet connection SMTP server?
If the SMTP server is under your control, is TLS correctly configured on it to modern standards? Do your server’s A/PTR records match the hostname it’s configured to use?
> You're suggesting that all email for YCombinator is down
No, I'm suggesting that their mail server is not reachable over unencrypted SMTP from random probably-residential IP addresses. This is fairly typical IME, and is why
> Using a post or comment is a terribly inefficient way to go about doing this [...] I'm exhausted of otherwise good-standing and high-karma users using posts and comments to bring things to their attention.
> maybe they didn't realize the Contact link exists or reaches the mods
is a non-sequitor. The point of comments as reporting medium is that they can be made via the same channel that was used to read the (offending or inaccurately-noted-as-offending) posts in the first place.
I emailed them a few minutes ago and they replied to confirm that they’re able to receive email.
While I respect that you may be demanding a certain standard of their mail server’s telnet responses that Google is refusing to meet, I imagine either that you are able to send them email and simply do not wish to, or that you have consciously chosen not to be able to email anyone hosting email with Google in deference to your objections at how Google operates. In any case, whatever the reason, I wish you the best of luck with your choices.
Given what the site represents and its current status, this isn't something to worry about. It'll get sorted.
"This project is a labor of love (or more accurately, a labor of like). Del.icio.us was founded by my friends in 2003, sold a whole bunch of times, and when it was about to get sold again to spammers in 2017, I took the opportunity to buy it back.
My current goal is to keep the site up and accessible as a museum of links past.
Right now, all you can really do here is log in and get your old data out."
His hypothesis if I understand it correctly, consciousness is the core part of existence so you would need to bring hardware into the world that not only is able to simulate the whole reality but to exist in it. Like property of physical reality and composition. That dude is a physicist believing in one single Universe, which probably also applies to most neuroscientists.
I'm not completely alien to the thought that someday we will get a digital companion that would be able to build a simulation of you in silico, but that simulation wouldn't be you.
This foundation sounds like complete science fiction, MRI really? We already have destructive approach in some countries - it's called euthanasia.
Abstract: How could machines learn as efficiently as humans and animals? How could machines learn to reason and plan? How could machines learn representations of percepts and action plans at multiple levels of abstraction, enabling them to reason, predict, and plan at multiple time horizons? This position paper proposes an architecture and training paradigms with which to construct autonomous intelligent agents. It combines concepts such as configurable predictive world model, behavior driven through intrinsic motivation, and hierarchical joint embedding architectures trained with self-supervised learning.
Link to pdf: https://openreview.net/pdf?id=BZ5a1r-kVsf