317070

18 ms·

241.
by 317070 3y ago
Generally yes and yes. But you will need _a lot_ more data and compute for reinforcement learning. So RL is usually too cumbersome for relatively little gain. Source: I'm a roboticist working with RL.
242.
by 317070 3y ago
Many European countries do not have a credit score system. Companies and banks are a lot less inclined to give you credit, but I don't see how that is a bad thing for society?
243.
by 317070 3y ago
Even if it isn't true. Disclosing inside information is illegal, _even if it is false and fabricated_, if it leads to personal gains.
244.
by 317070 3y ago
To be fair, compared to most EU countries, the level of accuracy and fairness of news in the Anglosphere is abysmal. If Die Zeit or Le Monde (Diplomatique) report something, in my opinion they tend to set the standard. In most (large enough
245.
by 317070 3y ago
I hope these numbers aren't correct. Only 3% for being bombed by ads? (And bombing the whole city with ads on buses as well?) I would gladly pay 20% and maybe even 50% more if they could stop with the ads.
246.
by 317070 3y ago
All right, I was already stuck in King's Cross next to this monstrocity. I was looking for a map, and I could only find these things that look like fake maps (because they are ads). Little did I know that unlike other campaigns with in
247.
by 317070 3y ago
Almost. The KL tells you how much additional information/entropy you get from a random sample of your distribution versus the target distribution. Here, the target distribution is defined as the unit gaussian and this is defined as the
248.
by 317070 3y ago
Not quite. We don't know the correct value for the ECS, and estimates for that value have been increasing in the last 5 years. Higher sensitivity means a higher temperature per extra carbon in the air.
249.
by 317070 3y ago
There is also this gem of a website, where you can explore a 3D CT-scan of a Bialetti moka pot, to see all its features and flaws: https://www.scanofthemonth.com/scans/coffee
250.
by 317070 3y ago
How would you enforce the three repetitions rule this way?
251.
Coronavirus can cause 100% mortality in hACE2 mice, risks spill over into humans(biorxiv.org)
2 points by 317070 3y ago|0 comments
252.
by 317070 3y ago
Yeah, that is what I'm implying here. I doubt Iran would already have known about Stuxnet by the time he was dead.
253.
by 317070 3y ago
In the original article [0] (in Dutch), it says: > Twee weken na zijn raadselachtige vertrek uit Iran overlijdt Van Sabben bij een ongeluk in Sharjah, nabij Dubai. Hij raakt op 16 januari 2009 met zijn motor van de weg, slaat over de kop
254.
by 317070 3y ago
Does ChatGPT beat that Turing test? No. But I'm pretty sure you could use the techniques behind ChatGPT to build an AI that can survive a two hour chat. Even ten hours is probably doable. There is just no obvious market to fund buildin
255.
by 317070 3y ago
But these computation graphs are similar every time. You would/should write these as a small graph with conditionals on large tensors, not as a fast changing graph of scalars. The latter will always be slower when you have an accelerat
256.
by 317070 3y ago
> 1. we test many computation graphs (graph is changing constantly) 2. many-many scalar operations with roughly 10k—100k nodes in each graph 3. every graph should be compiled and ran around 10k times both forward and backward Now I&#x2
257.
by 317070 3y ago
... because we have been working to this moment for often 10+ years. People that were working on this stuff for a long time were not in it for money. It used to be a more artsy/philosophical field, and the money people moved on a long
258.
by 317070 3y ago
> our awareness of the galactic bodies itself is somehow influencing them, as in the double-slit experiment In the double slit experiment, the awareness is not influencing the outcome. The act of measuring is. Pretty sure the act of mea
259.
by 317070 3y ago
Exactly, I think we 100% agree.
260.
by 317070 3y ago
Given that anyone can still doodle, sing and create, even in the vicinity of generative AI, makes me think the author was thinking of mass media when he wrote the piece.
261.
by 317070 3y ago
> Imagine a world where everything is “AI generated”, everything including books, songs, music, movies, etc. One can imagine the irony in humans consuming and enjoying such content - the very same humans who have the potential to create
262.
by 317070 3y ago
It's as if proteins contain a lot of amino acids.
263.
by 317070 3y ago
Wait, we know the underlying rules, we have an explanation. You can read all the coefficients. We don't understand the explanation, though it is correct. Not sure if the problem here is with the capabilities of the examinee or with the
264.
by 317070 3y ago
How many pounds of veggies would a person eat per year? How many would they need to sustain themselves?
265.
by 317070 3y ago
Sounds like the movies "Don't look up" and "Melancholia" Not much really happens, all everybody knows is that doom is imminent. The movies are mainly about how the different characters handle this situation. That sa
266.
Who Is the Greatest Economist of All Time? A Generative Book(econgoat.ai)
1 points by 317070 3y ago|0 comments
267.
by 317070 3y ago
Frankly, I don't see how discussing the expense side is useful in a discussion on how we distribute the income side of the government over the population.
268.
by 317070 3y ago
It is indeed saying that, you can't model secondary effects and they are ignored in this analysis. If you believe trying to have the rich pay more of the government budget is only going to lead to them paying less, no other assumptions
269.
by 317070 3y ago
You see, the reason you pay so much money on your tax income, is _because_ we are not or only barely taxing wealth income. Cfr. Piketty, governments balance sheets would stay the same (on a global scale) if you would drop the income tax on
270.
by 317070 3y ago
> You can't represent 1/50 exactly in binary You can though? You represent the encoder state in fractions of 50. In practice, it means the encoder state will not use all possible binary strings. But you don't need to trade

More ›