Changing that 0.1 to 1 would make the bullet move 10x faster. It'd be neat if you could have multiple bullets on the screen at the same time, but then the developer would need to maintain an array of bullets and update all of their .y values.
In this case, a new bullet is created when the .y value gets to zero, so there isn't an easy way to increase the firing rate except to increase the speed of the bullet.
By "easy" I mean, can be done in under 5 keystrokes to the code and without introducing potential bugs. Adding an array of bullets moving at the same rate, but able to be fired with less time delay between each bullet, would require at least an order of magnitude or two more than that amount of effort and also a not insignificant amount of testing.
It's hard to verbalize this, most of it is "intuition" but I think it boils down to "supervised learning is BS."
Humans are smarter than computers. How can a human teach a computer how to do something when the human itself can't teach another human that something?
We haven't solved that problem. The snake is eating its tail.
You can't teach a human how to do something when the methodology to do that is the student trying something and the teacher saying "Yes" or "No".
Well.... why? Why is it yes or why is it no? What is the difference between what the human or the computer, or in general, the student, did and what is good or correct? And then you still have to define "good" and many times that means waiting, in the case of the PDF linked to above, perhaps many years to determine if the employee the AI picked, turned out to be a good employee or not.
And how do you determine that? How do you know if an employee is good or not? We haven't even figured that out yet.
How can we create an AI to pick good employees if human beings don't know how to do that?
Supervised learning isn't going to solve any problem, if that problem isn't solved or perhaps even solvable at all.
In other words, over the years, my heuristic has turned into, "Has a human being solved this problem?" If not, then AI software that claims to is BS.
Supervised learning in machine learning is nothing remotely like a human teaching anyone anything. It's a very clear mathematical formulation of what the objective is and how the algorithm can improve itself against that objective.
The closest analogy for humans would be to define a metric and ask a human to figure out how to maximize that metric. That's something we're often pretty good at doing, often in ways that the person defining the metric didn't actually want us to use.
> Supervised learning in machine learning is nothing remotely like a human teaching anyone anything.
I disagree, I think it's exactly the same. As an example, a human teaching a human how to use an orbital sander to smooth out the rough grain of a piece of wood.
The teacher sees the student bearing down really hard with the sander and hears the RPM's of the sander declining as measured by the frequency of the sound.
The teacher would help the student improve by saying, "Decrease pressure such that you maximize the RPM's of the sander. Let the velocity of the sander do the work, not the pressure from your hand."
That's a good application of supervised learning. Hiring the right candidate for your company is not.
But that's not at all how "supervised learning" works. You would do something like have a thousand sanded pieces of wood and columns of attributes of the sanding parameters that were used, and have a human label the wood pieces that meet the spec. Then you solve for the parameters that were likely to generate those acceptable results. ML is brute force compared with the heuristics that human learning can apply. And ML never* gives you results that can be generalized with simple rules.
Machine learning really almost nothing in common with most types of human learning. The only type of learning that has similarities is associative learning (think Pavlovs dogs studies).
The human learning situation you describe works quite differently, though: The student sees either the device alone or the teacher using the device to demonstrate its functionality. This is the moment most of the actual learning happens: The student creates internal concepts of the device and its interactions with the surroundings. As a result the student can immediately use the decive more or less correctly. What's left is just some finetuning of parameters like movement vectors, movement speed, applied pressure etc.
If the student would work like ML, it would: hold the device in random ways, like on the cord, the disc, the actual grip. After a bunch of right/wrong responses she would settle on using the grip mostly. Then (or in parallel) the student would try out random surfaces to use the device on: the own hand (wrong), the face of the teacher (wrong), the wall (wrong), the wood (right), the table (wrong) etc. After a bunch of retries she would settle on using the device on the wood mostly.
It's easy to overlook the actual cognitive accomplishments
of us humans in menial tasks like this one because most of it happens unconsciously. It's not the "I" that is creating the cognitive concepts.
> You can't teach a human how to do something when the methodology to do that is the student trying something and the teacher saying "Yes" or "No".
Strangely, I recently had to complete a cognitive test that was essentially that process. I was given a series of pages, each of which had a number of shapes and a multiple choice answer. I was told whether I chose the correct answer, then the page was flipped to the next problem. The heuristic for the correct answer was changed at intervals during the test, without any warning from the tester. I'm told I did OK.
You're touching on the "difficulty" in verbalizing it. I see what you mean, because you did learn that the heuristic was changing with just a yes or no. I said you can't teach that way, but you clearly learned that way, so I wasn't exactly correct, but I'm not practically wrong either still I don't think.
I wonder, how would an AI perform on the same test.
What is the mathematical minimum number of questions on such a test, subsequent to the heuristic change, that could guarantee that new heuristic has been learned?
I'm curious about the test. Did it have a name? What were they testing you for?
> I wonder, how would an AI perform on the same test.
This situation is called Multi-armed Bandit. In this setup you have a number of actions at your disposal and need to maximise rewards by selecting the most efficient actions. But the results are stochastic and the player doesn't know which action is best. They need to 'spend' some time trying out various actions but then focus on those that work better. In a variant of this problem, the rewards associated to actions are also changing in time. It's a very well studied problem, a form of simple reinforcement learning.
Doesn’t it depend on what you mean by guarantee? The test can’t get 100% certainty, since theoretically you could be flipping a coin each time and miraculously getting it right, for 1000 times in a row. The chance of that is minuscule (1/2^1000), but it’s nonzero. So we’d have to define a cutoff point for guaranteed. The one used generally in many sciences is 1/20 chance (p = 0.05), so that seems like a plausible one, and with that cutoff, I think you’d need five questions passed in a row (1/2^5 = 1/32). Generally, if you want a chance of p, you need log2(1/p) questions in a row passed correctly. However, that only works if your only options are random guessing and having learned the heuristic. If you sorta know the heuristic (eg. right 2/3 of the time), then you’d get the 5 questions right ~13% ((2/3)^5) of the time, which isn’t inside the p = 0.05 range. So you also need to define a range around your heuristic, like knowing it X of the time. Then you’d need log(1/p)/log(1/X) questions. For example, if you wanted to be the same as the heuristic 19/20 times and you wanted to pass the p = 0.05 threshold, you’d need log(1/0.05)/log(1/(19/20)) ~= 59 questions.
I'm sure the test was a standard with a name, but I was never told. It was a small part of a 3 hour ordeal, evaluating my healing progress since suffering a brain injury in March.
I would agree that it's a very inefficient way of teaching something. It gave me an unexpected insight into machine learning though.
I'm sure the test was designed so that picking the same answer each time or picking one at random would result in a fail.
I agree and rather than post a sibling response, I'll add that I think it's necessary today, simply because we don't have AGI, yet. And also point out that we are talking about determining if AI is snake oil or not. There may be some scenarios where we can teach a computer to do something we can't teach a human to do, I can't think of any off the top of my head, but if we can't, then I'm going to be super doubtful that an AI software can do it better than a human, if at all.
AGI, in the singularity sense, will be solving problems before we even identify them as problems. Experts in a field can do this for the layman already and I think it's possible. Some don't. I do.
It'll be super interesting when it flips! When the student becomes the master and we, as a species, start learning from the computer. You can kind of get a sense of this from the Deep Mind founder's presentation on their AI learning how to play the old Atari game Breakout. He says when their engineers watched the computer play the game, it had developed techniques the engineers who wrote the program hadn't even thought of.
Even still, the engineers could teach another human how to play Breakout, so yes, I do believe they did in fact create a software to play Breakout better than they could.
Same for AlphaGo, but it only works when you have access to cheap simulation (breakout being a game easy to run, Go being just a board). It doesn't work as well in situations where you don't see the full state of the system, or where there is randomness.
This simply isn't true. We know your intuition here is mistaken, as we have plenty of counter-examples.
The best chess AIs can beat any human chess player. They use techniques that were never taught to them by a human.
Another example: a machine-learning-driven computer-vision system predicting the sex of a person based on an image of their iris. No human can do this. [0]
You and the GP should have both stopped before saying "It's theft." It's not theft, not even close.
You gave them the video and they said they do not want the video -- for whatever reason. You don't have a right to upload content to YouTube. You do have a right to your property, your video and music you've created. They didn't take it from you. They aren't making money on it and not paying you. You don't have a right to force YouTube to display and distribute that content on their website.
Imagine if Columbia records was forced to distribute any and all music sent to it on a demo tape. Columbia saying, "No, we won't distribute this" is not theft.
Columbia would be within their rights to say, "Your music sounds too much like Taylor Swift to us. Sorry. It's too big of a risk for us to distribute it."
That's exactly what YouTube is doing. They are saying, your music is too much of a risk for us to distribute based on our algorithms.
You can still go to Vimeo or soundcloud or build your own website to distribute your music, just like the aforementioned artist that sounds too much like Taylor Swift could go to Arista or upload it to youtube or or soundcloud.
Seems a bit more like you sending your tape to Colombia and them deciding to distribute it but give all proceeds to some other guy with no relation to your music.
ed: You’ve edited your comment but I was referring more to the GP case where they say
> And yet, they have claimed copyright on my music and monitized it.
That is exactly Youtube making money on the video and not paying the owner
That's not what I understood when I read it the GP saying:
>And yet, they have claimed copyright on my music and monitized it.
By "they," I think he meant: ASCAP, ICE_CS, not YouTube. The problem is that ASCAP, ICE_CS has claimed copyright to something, which they are right to do and that is their performance of the public ___domain work. Since it sounds like the same work that the GP uploaded, the GP's music was locked.
I am no fan of YouTube's profiting off copyrighted material, I think it's unethical, and in this instance I do not believe that is what they are doing. I believe they have written algorithms to prevent copyright infringement and it is over-fitting on the GP's works.
The GP said they are "claiming copyright on my music" but I don't see any evidence for that. I do see evidence that ASCAP, ICE_CS claim copyright on their performance of the same music as the GP.
Yes, the solution is to have a person in there to be smarter than the AI, but YouTube has said it's not worth it to them to do that and so I believe that is their right.
In this case, the GP's claim is that ASCAP, ICE_CS is doing something wrong, but I don't think that is the case. I think everyone here is doing something right and a bad thing has come from that.
My understanding is that Youtube will allow the supposed content owner to monetise your video. They can choose to let you keep it up as long as they get the advertising revenue from it.
I did edit to elaborate where I thought there was ambiguity.
Yes, I agree, if ASCAP, ICE_CS laid claim to the GP's work, then that needs to be fixed. I would suggest any money paid to ASCAP, ICE_CS for that video be returned to the GP, with interest paid by ASCAP, ICE_CS and perhaps some punitive damages so they don't make the same "mistake" again.
To be fair, or give the benefit of the doubt, this could be a Hanlon's Razor scenario where ASCAP, ICE_CS just doesn't understand how to do it right. Or they could be too heavy handed in which case the punitive damages should soften their hand.
I came here to say the same thing as the GP. I don't understand why some words are red or green.
For example, you can type in non-brand words as well. I typed in "houses" and the word "homeless" came up in green!
With a brand, facebook, I got this word "amiriteguyze" in red and clicking on it
Negative
11/19/2019, 12:13:31 PM
facebook is bad amiriteguyze?!?!?!?
Why is that even a word that would show up in the word cloud? I can't imagine it was entered a bunch of times. I can't intuit any correlation between the colors, sizes, or words themselves that show up in the clouds.
The algorithm will try to give more importance to words which appear rarely and are only used with the chosen brandname (similar to TF-IDF). This is why sometimes weird words can surface to the wordcloud, especially when the sample size of messages is small.
To prevent those words from appearing, I was thinking to implement some dictionary-check to only allow for meaningful words. However this approach also have drawback as you restrict people's words and can miss important new concepts.
> What’s the “magic” behind this? Executives are coy, but biotechnology expert Oded Shoseyov, a Hebrew University professor who has consulted for UBQ, says melting plastics and waste creates a homogeneous substance strengthened by fibers in the organic ingredients.
The article then goes on to describe a typical waste recycling facility plus:
> a multi-chamber reactor that sits behind a closed sliding door to block prying eyes. Temperatures up to 400 degrees break down the organic matter into its core elements, and then it and the plastics are re-engineered into a matrix through chemical and physical reactions that UBQ keeps secret.
I fear there is really no story here. There is a lot of energy being put into this system. Something comes out, yes, but is it worth it? How is it better than a garbage incinerating facility?
There is no evidence presented that the material output is as good as the material input. Or that the energy required is less than traditional recycling methods.
This is way too early and way too long of an article to spend a lot of time on at this point.
Maybe we can use this and stop pulling new petroleum into the plastics cycle for most products. That would let us leave more of it in the ground. And this also lets us prevent the gases from the food, much of it generated using petroleum products, from entering the atmosphere, effectively sequestering carbon into the products.
If all the system really needs is heat and mechanical energy we can generate that in spades from sunlight. In fact, the sun is far more efficient at directly heating things up than it is at being turned into electricity. So we could actually have a recycling plant that for certain kinds of materials is far greener than throwing the materials away.
First we need to demonstrate that the process works, and that we can spend the energy to make something useful. Then we figure out if we can source the energy from somewhere green.
The only real questions I have are what are the chemical inputs, and where do they come from? And can those chemical inputs be generated directly using heat or do they require electricity?
It's interesting because it suggests that we can build a recycling machine that takes garbage and heat and outputs usable materials.
It suggests that's possible, but we know that's possible already.
The hard part of plastics recycling which is already being done at various scales is isolating the plastic. How do you get foodstuffs out of plastic containers? How do you separate the components of waste, basically?
It's like cleaning metal by burning off everything that burns until all you have is metal. You're just moving the waste from one place to another.
In this case, the gasses are getting into the atmosphere just like they would at a traditional incinerating waste facility and there are lots of those already. They are not a good solution for the long term.
When people are being "coy" about something as the article indicates, there's usually a reason. In this case, I think it's because the solution solves the problem, but not better than existing technology and they don't want that to be revealed to the open market that would tear it apart.
To add: The music needs to be loud enough to drown out every other aural stimulus. You want to hear the music and only the music, not everyone's side conversations, not your clothes rustling around while you're dancing, not the sound of people's hard healed shoes on the dance floor.
HN is a marketing channel for startups who might want to pitch ycombinator, then market the startups they pick, then find employees for those startups, then repeat.
The value isn't in selling ads.
I see this whole line of reasoning that thousands of people isn't enough to be newsworthy a symptom of valleythink where all that is valued is growth.
There are other factors that indicate success and the community for the most part has lost sight of those for the tunnel vision of "must get funding!"
> Once you start distributing your HTML5 game on PC distribution platforms like Steam or GOG, you need an actual executable that people can download to play the game offline as well.
Can someone elaborate on this? Why is this needed? Is it a market penetration decision? Does steam not allow HTML5 games?
Due to virus issues, etc, I would prefer playing an HTML5 game in the browser. I don't want to install something that could contain viruses or be a vector or increase my attack surface.
I'm missing something, but not sure what it is. I don't use Steam or install games on my computers, but I do play HTML5 games.
When customers of a PC distribution platform spend money on your game they want to be able to own it by downloading it and having it live on their computer and work offline as well. For this you need to deliver them something that is completely packaged and contains everything to run the game.. so even if that means basically shipping a full browser with your HTML5 game.
We also have a web version which can be played online via the regular browser.
Then you'll get a flood of support tickets from people whose default browser doesn't run the game quite right. People downloading a game on Steam couldn't care less if it weighs an extra 100MB because it includes a copy of Chromium.
Once you are charging money for the game on a distribution network, you probably aren't providing the full game on a standalone site anymore. Maybe you have a small demo.
That's already a possibility if you're using Steamworks (which most Steam games - or at least nearly all I've played - are). Granted, the breakages from a Steamworks-related update are likely far less critical (unless your game critically depends on achievements, cloud saves, instant screenshots, or the Steam Overlay, which would be weird), but it ain't like this would be entirely unprecedented.
Breakage is also already a possibility for games with Linux ports that rely on the Steam-provided runtime environment (I don't know if this applies to Windows or macOS, since I haven't used the former with Steam in years, and haven't used the latter with Steam ever).
I always wonder about claims like this where one nation says a ship was within their borders and the other nation says it was in international waters. This USS Pueblo is one incident. Israel attacked a Turkish ship in international waters. Iran attacked oil tankers in international waters.
All these incidents ... I can't help but think we never grow up as human beings all the way until becoming international super powers.
Or like the US and Russia always intercepting each other's aircraft. When reported in the media there's often a lot of bellyaching, i.e. "we were just minding our own business flying in international airspace, and then they showed up and intercepted, and they were so rude!" Like, maybe if you hadn't been flying your bombers or your spy planes 13 nautical miles from their coast they would be a bit friendlier. Nobody buys your "poor me" bullshit when it's clearly an example of a defensive response to an aggressive provocation.
My favorite in particular in this line of "geopolitical I'm not touching you" is what we call "freedom of navigation exercises". The US runs through the South China Sea on them all the time.
Basically, China says it's their territory, the US says it is international waters. And so we "prove" it by regularly traveling through it with US warships for no other purpose then to prove we can.
> China says it's their territory, the US says it is international waters. And so we "prove" it by regularly traveling through it with US warships for no other purpose then to prove we can.
Suppose the US claimed that the entire Gulf of Mexico was US territory. Would people be as sympathetic to that as they appear to be to China claiming that the entire South China Sea is Chinese territory? (Note that the South China Sea is more than twice as large as the Gulf of Mexico, and less of its coastline is part of China as compared with the amount of Gulf of Mexico coastline that is part of the US.)
> Since China is a long way from the Gulf of Mexico, and has no need to sail there, I'm confident that China couldn't care less about this.
First, many Chinese flag ships sail through the Gulf of Mexico, so I think China does care.
Second, even if China didn't care, many other countries do. And the US, unlike China, respects the rights of other countries to freedom of the seas.
> The South China Sea is very important to China, and very far away from USA -- why does USA care so much about it?
First, because many US flag ships sail through the South China Sea. We trade with many countries other than China that have coastlines on that sea. You know that, right?
Second, because, just as with the Gulf of Mexico, many other countries have ships that sail through the South China Sea, and there are internationally agreed rules about international waters and freedom of the seas, which, as noted above, China does not respect the way the US does. The only way to enforce such rules is for countries that uphold them to take action against countries that do not.
> Perhaps you could also comment of USA claims over the Northern Sea Route (which lies within Russia’s Exclusive Economic Zone) ?
This case is an excellent illustration of the proper way to negotiate rules regarding freedom of the seas: to notify other countries of something you would like to see happen, and then have a diplomatic dialogue with them. As opposed to, oh, say, just asserting that a huge chunk of ocean is your territorial waters, in contravention of international law and treaties.
Also, since you mentioned Exclusive Economic Zones, you will note that the US is not saying it wants to exploit resources in this zone; it is only saying it wants it to be a transport corridor. Which, according to current international agreements, it already is everywhere outside 12 nautical miles from Russia's coast.
China sends in soldiers as police, does false flag attacks, hires the triads, tear gasses 90% of the city, kills protesters on the streets, and arrests protesters and fakes their suicides. I'd much rather deal with US police.
This is one of those situations where I have to hold two conflicting ideas in my head at the same time and am reminded of the difficulty of doing that.
You both make good points. China is doing a lot of bad stuff here, but it is their land and their right to exert their power over it how they see fit.
If China tried to tell another sovereign nation how to handle their internal disputes, they'd be just as wrong as the west is in this.
This is really something China needs to handle. It's not the west's place to butt in.
At the risk of whatabouting it, most of all the tactics you're saying China is using in Hong Kong have been used by western governments as well. It's not right.
You know, China gave away Hong Kong to stop the bombing from the west and now China is trying to wrestle back control of it... kind of still, in a way, from the west.
It was the west that caused this problem in Hong Kong. It's time the west stopped meddling in it and I think that'd be best.
You essentially said we're not grown up because we don't live in some kind of magical world where nations don't have competing interests that manifest in these sorts of displays. Competition is inherent to life, and, if anything, growing up means accepting that.
If any browser can be a camera, why do you require me to register to try it out? I'm not ready to give you my personal details. You say below privacy is your primary concern, but you require us to give up our privacy just to try it out.
I guess they could do without the registration part if the demo only involved streaming your own video back to the same device. But the primary use case seems to be streaming from one private device to another (not broadcasting to the world), and for that they need a way to determine which devices are allowed to access which streams.
In most web frameworks, accepting signups and associating sessions with user accounts is the safest and easiest way to achieve this.
In this case, a new bullet is created when the .y value gets to zero, so there isn't an easy way to increase the firing rate except to increase the speed of the bullet.
By "easy" I mean, can be done in under 5 keystrokes to the code and without introducing potential bugs. Adding an array of bullets moving at the same rate, but able to be fired with less time delay between each bullet, would require at least an order of magnitude or two more than that amount of effort and also a not insignificant amount of testing.