Glad to hear you also get it now.
To all the other shit replies: If you invested in Google or any other fundamental tech startup at the time you would have made at least a x10. In Google's case a x30. Ever heard of angel investing? They also return x10 quite frequently, sometimes x100 or x1000. The only difference with ICOs and traditional investments is that we WILL get screwed over eventually with our useless ERC-20 tokens, myself included.
Also don't you dare use the pro-government strawman argument. If you want real decentralized tokens then these ICOs should be launched as proper DAOs. DAOs are fully decentralized and embrace the spirit of the blockchain, whilst conferring actual ownership of the organization with the tokens. Registered companies that are motivated by profit that issue us shitty utility tokens have nothing at all to do with decentralization or any anti-government movement. In fact they will comply with governments if they're forced to. DAOs don't have this issue. The majority of ICOs are not DAOs. Think of better arguments if you don't agree.
From mathematical point of view, the fact the you don't own any equity in any of those companies is just another factor that should be considered in expected value equations. So, if you know that there's a non-negligible probability that you might get robbed and you include that probability into your profitability assessment and you still think that investing is smart and profitable, go for it. This is somewhat similar to a situation where professional poker player decides to play on shady unlicensed poker room: he carefully compares the probability of getting robbed and/or cheated to the "juicyness" of games that are being played there and decides accordingly.
One more advantage of ICO investing over traditional security investing is that you don't have to go though a lot of regulatory garbage. Though with all those KYC rules that are implemented nowadays, this is becoming less and less true.
Is your head really that far in the clouds with investment? It's common knowledge in investment that crowd psychology is what matters, not efficient market hypothesis that you read about in the textbook.