It's not really, but the difference is that I'm limited by the supply of BTC, and it requires that I actually have the money to make the 'bet' at the start. That restricts the size of the spot market.
If I'm buying futures I can enter into a contract that says "I'll buy a contract for 1BTC that says BTC is going to go from $88.5k to $98.5k in 1 year." I don't actually hand over any money. In a year's time, if BTC is now $100k the person who agreed on the contract gives me $10k. If it doesn't go up then I owe the seller $10k. The futures contract is settled in cash - no BTC is involved.
Right now though, I don't have a $88.5k to spend on BTC, so the spot market isn't an option. I probably could find $10k in a year's time so a bet on a BTC future might be viable. The actual derivative 'value' isn't real though. The only money changing hands is the delta of the change in value when the contract is settled.
(Caveat: I am a total noob at finance stuff so this could be quite wrong. One of the many reasons I will not be buying that futures contract. :) )
It's very wrong. Futures contracts on traditional exchanges have no counterparty risk and require the deposit of a significant amount of upfront capital as collateral. If the spot price of the underlying moves in either direction, debits or credits are made to and from each margin account and if you don't have the money to cover a margin call, the contract gets closed.
Margin requirements for trading spot are zero, though initial capital requirements are obviously, well, whatever spot is.
Futures contracts aren't just pieces of paper traded between people, they are actual promises to pay for physical delivery of the underlying.
It's not surprising to me that crypto people consider them nothing more than leveraged gambling slips but that's really not how one should think about them. Personally I think crypto needs far heavier regulation than it gets.
Derivatives can be structured in a time-constrained manner that requires them to go up/down in a specific time window, thus amplifying the gains/losses. Also there's generally no way to short an asset without borrowing them with a contract to pay them back (which requires timing the market move and paying rent on the asset). This is something that options contracts solved.
Its hard for me to consider owning the underlying asset as gambling compared to owning paper bets on the future value. In the former you are owning it today, in the later you are betting only on what it will cost to own later.
Why hasn't Java added an option for automatic heap sizing for the desktop application use case (or any use case, for that matter)? Seems like a no-brainer thing to add, so I guess there are Reasons(tm).
It has. Modern JVMs can use as much memory as they "need" to, but for some reason JetBrains don't use this feature. I think after they upgrade to Java 17 they'll get it. But this has really killed their brand for a long time. 99% of the time people complain about their IDEs being slow it's because the app is GC thrashing and burning CPU rather than increase heap size.
Good article, but there is a frequent misconception of Symbolic AI as "manual creation of lots of rules". This was true for early approaches, such as expert systems in the 70s/80s. Symbolic AI just means, well, AI with symbols, and there are many approaches (e.g., in neuro-symbolic AI or using probabilistic inductive logic programming) where symbolic representations are emergent / learned from data using machine learning approaches and can be uncertain/probabilistic.
In the linked talk "From System 1 Deep Learning to System 2 Deep Learning" by Yoshua Bengio, the speaker first criticizes Symbolic AI only to re-invent concepts from Symbolic AI later (e.g., "high level semantic variables", "shared 'rules' across arguments"), which is rather silly given that some Symbolic AI approaches are well capable of learning symbols, rules etc bottom-up - which is not fundamentally different from learning low-dimensional vector representations or "generalizations" in linguistics.
Thanks for the reply. I agree there might be a lot of different approach to to symbolic AI. Probabilistic inductive logic programming sounds interesting but I am not familiar with it. Maybe I should check it out later. I am wondering if it has capability to learning from data.
While the syntax of Answer Set Programming is similar to Prolog, inference in ASP is closer to SAT solving, and ASP solvers are typically extensions of SAT solvers.
