All sites
Data API
Hackernews Data API
Structured JSON data, on demand, with anti-bot bypass built in.
9endpoints
Example (captured live) Open playground
facebook.com
ExampleJSON responseclick the page to trace a field
{
"id": 21530860,
"title": "John Carmack: I’m going to work on artificial general intelligence",
"url": "https://www.facebook.com/100006735798590/posts/2547632585471243/",
"by": "jbredeche",
"score": 1574,
"comments_count": 889,
"time": "2019-11-13T23:17:23+00:00",
"type": "story",
"text": null,
"hn_url": "https://news.ycombinator.com/item?id=21530860",
"domain": "facebook.com",
"comments": [
{
"id": 21532491,
"by": "Animats",
"time": "2019-11-14T03:25:10+00:00",
"text": "This is encouraging. If you're going to work on artificial general intelligence, a reasonable context in which to work on it is game NPCs. They have to operate in a world, interact with others, survive, and accomplish goals. Simulator technology is now good enough that you can do quite realistic worlds. Imagine The Sims, with a lot more internal smarts and real physics, as a base for work.<p>Robotics has the same issues, but you spend all your time fussing with the mechanical machinery. Carmack is a game developer; he can easily connect whatever he's doing to some kind of game engine.<p>(Back in the 1990s, I was headed in that direction, got stuck because physics engines were no good, made some progress on physics engines, and sold off that technology. Never got back to the AI part. I'd been headed in a direction we now think is a dead end, anyway. I was trying to use adaptive model-based control as a form of machine learning. You observe a black box's inputs and outputs and try to predict the black box. The internal model has delays, multipliers, integrators, and such. All of these have tuning parameters. You try to guess at the internal model, tune it, see what it gets wrong, try some permutations of the model, keep the winners, dump the losers, repeat. It turns out that the road to machine learning is a huge number of dumb nodes, not a small number of complicated ones. Oh well.)"
},
{
"id": 21533661,
"by": "Jaruzel",
"time": "2019-11-14T08:04:32+00:00",
"text": "Carmack's post in full:<p><i>Starting this week, I’m moving to a "Consulting CTO” position with Oculus.<p>I will still have a voice in the development work, but it will only be consuming a modest slice of my time.<p>As for what I am going to be doing with the rest of my time: When I think back over everything I have done across games, aerospace, and VR, I have always felt that I had at least a vague “line of sight” to the solutions, even if they were unconventional or unproven. I have sometimes wondered how I would fare with a problem where the solution really isn’t in sight. I decided that I should give it a try before I get too old.<p>I’m going to work on artificial general intelligence (AGI).<p>I think it is possible, enormously valuable, and that I have a non-negligible chance of making a difference there, so by a Pascal’s Mugging sort of logic, I should be working on it.<p>For the time being at least, I am going to be going about it “Victorian Gentleman Scientist” style, pursuing my inquiries from home, and drafting my son into the work.<p>Runner up for next project was cost effective nuclear fission reactors, which wouldn’t have been as suitable for that style of work. </i><p>--<p>We're at 500 comments at the time of posting this, and no-ones pasted his post in full to save us having to visit Facebook..."
},
{
"id": 21531092,
"by": "pyentropy",
"time": "2019-11-13T23:44:18+00:00",
"text": "Progress in AI is due to data and computational power advances. I wonder what kind of advances are needed for AGI.<p>1. Biological brains are non-differentiable spiking networks much more complicated than backpropagated ANNs.<p>2. Ion channels may or may not be affected by quantum effects.<p>3. The search space is huge (but organisms aren't optimal and natural selection is probably local search)<p>4. If it took ~3.8b years to get from cells to humans, how do we fast-forward:<p>* brain mapping (replicating the biological "architecture")<p>* gene editing on animal models to build tissues and/or brains that can be interfaced (and if such interface could exist how do we prevent someone from trying to use human slaves as computers? Using which tissues for computation is torture?)<p>* simulation with computational models outside of ECT (quantum computers or some new physics phenomenon)<p>Note: those 3.8b years are from a cell to human. We haven't built anything remotely similar to a cell. And I'm not claiming that an AGI system will need cells or spiking nets, most likely a lot of those are redundant. But the entropy and complexity of biological systems is huge and even rodents can outperform state of the art models at general tasks.<p>IMHO, the quickest path to AGI would be to focus on climate change and making academia more appealing."
},
{
"id": 21531255,
"by": "carlosdp",
"time": "2019-11-14T00:04:48+00:00",
"text": "This doesn't surprise me at all. He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). (edit: I'm not claiming he's a field expert in a week guys, just that he can probably learn the basics pretty fast, especially given ML tech shares many base maths with graphics)<p>As recent as his last Oculus Connect keynote, he extolled his frustration with having to do the sort of "managing up" of constantly having to convince others of a technical path he sees as critical. He's clearly the type that is happiest when he's deep in a technical problem rather than bureaucracy, and he likes moving fast.<p>On top of that, he likes sharing with the community with talks and such, and ever since going under the FB umbrella, he's had to clear everything he says in public with Facebook PR, which clearly annoyed him.<p>He's hungry for a new hard challenge. VR isn't really it right now since it's more hardware-bound by the need for hard-core optical research than software right now. With the Quest, he (in my opinion) solidified VR's path to mobile standalones. It's time to try his hand at another magic trick while he's on his game.<p>John's the very definition of a world-class, tried and true engineer/scientist. He's shown time and time again the ability to dive into a field and become an expert very quickly (he went from making video games to literally building space rockets for a good bit before inventing the modern VR field with Palmer).<p>If there's anyone I'd trust to both be able to dive into AGI quickly and do it the right(tm) way, it's John Carmack."
},
{
"id": 21532524,
"by": "xamuel",
"time": "2019-11-14T03:31:41+00:00",
"text": "I've been dabbling in AGI and it seems like the field has a lot of low-hanging fruit. I'll bet Carmack can offer some significant contributions.<p>I'll take an opportunity to plug a paper I recently published on comparing relative intelligence. The punchline will illuminate the low-hangingness of the fruit in this field.<p>Suppose X and Y are AGIs and you want to know which is more intelligent. For any interactive reward-giving environment E, you could place X into E and see how much reward X gets; likewise for Y. If X gets more reward, you can consider that as evidence of X being more intelligent. But there are many environments, and X might do better in some, Y in others. How can you combine those pieces of evidence into a final judgment?<p>The epiphany I had (obvious in hindsight) is that the above situation is actually an <i>election</i> in disguise. The voters are interactive reward-giving environments, voting (via their rewards) in an intelligence contest between different AGIs. This allows us to import centuries of research on voting and elections! In particular, by using theorems about elections published in the 1970s, I was able to provide an elegant notion of relative intelligence.<p>The notion I provided is elegant enough that some theorems can even be proved with it, for example, formalizations of the idea that "higher-intelligence team-members make higher-intelligence teams". Which emphasizes the low-fruit-hanginess of the field: as obvious as that idea seems, apparently no-one was able to prove it with previous formal intelligence measures, probably because those previous intelligence measures were too complicated to reason about!<p>Here's the paper: <a href="https://philpapers.org/archive/ALEIVU.pdf" rel="nofollow">https://philpapers.org/archive/ALEIVU.pdf</a>"
},
{
"id": 21531021,
"by": "phillco",
"time": "2019-11-13T23:36:59+00:00",
"text": "This is a very artfully written statement. It avoids the most more devastating headline of “John Carmack Stepping Down”. (Yes, going part-time isn’t the same as leaving entirely, but still)."
},
{
"id": 21531870,
"by": "nwsm",
"time": "2019-11-14T01:26:21+00:00",
"text": "What a joke. Carmack is going to sit at home and solve what teams of scientists can't do in decades.<p>I'm complaining less about Carmack wanting to spend his time doing this and more about the comments here acting like he is some 10000x research scientist."
},
{
"id": 21531085,
"by": "yaseer",
"time": "2019-11-13T23:43:22+00:00",
"text": "John Carmack seems like a guy that would've made fundamental contributions to science, had he been born a century before.<p>Computer science now occupies the place physics once did, in its impact on moving the world forward.<p>Best of luck to him, I look forward to seeing what he produces!"
},
{
"id": 21530903,
"by": "hans1729",
"time": "2019-11-13T23:22:43+00:00",
"text": "Heh, from the comments:<p>><i>Congratulations on the new project, and may your hubris not doom us all.</i>"
},
{
"id": 21532485,
"by": "leesec",
"time": "2019-11-14T03:22:56+00:00",
"text": "Wow, there are a lot of people in this thread arguing whether or not John Carmack has the right skills to help AGI, or about the specifics of his knowledge.<p>Do you all realize you're arguing about nothing?<p>Good for him for doing something he seems excited about. Maybe we should all stop gossiping and go do something we're excited about too."
}
]
}
GET
/v1/hackernews/items/{id}1 credit / run VerifiedItem
A single story, comment, job or poll by id. For a story, also returns its first top-level comments.
Parameters
| Name | Type | Req | Description |
|---|---|---|---|
| id | string | yes | Item id. |
| Advanced filters | |||
| comment_limit | integer | no | Top-level comments to include for a story (0-50). (default: 10) |