There are so many things wrong in this article, I just can’t... https://cointelegraph.com/news/switching-bitcoin-to-100-asic-mining-could-increase-security-2000x/amp?__twitter_impression=true

1. There are plenty of FPGA and GPU coins where the majority hashpower sits on one entity. They’re not 51% attacked. Why? Because it would bloody invalidate the chain, and make the coin worthless.

2. Attacking a coin is mostly solely done as a competitive move from other coin projects nowadays (yes, this happens in the alt coin world). It shuts down their liquidity by removing them from an exchange, as well as tanks pricing.

3. The benefit you get from “double spending” is mostly meaningless when you can’t cash out these coins at an exchange. It’s all warfare. 51% attacks happen in the “deep” shitcoin space regularly - solely to harm competing coins.

4. These 51% attacks happen with SHA256d ASICS too. Why? Because golly gee, the hardware isn’t as useless as our researchers would lead you to believe! When RVN forked, the X16Rv2 ASIC commanded a 215% price premium because of the fork coin, RVC.

5. It continued to command that premium until an X20 ASIC replaced it. Even though, arguably, it had been invalidated.

As long as the HW exists, people will create reasons (new coins) for its existence. It takes <35 min to fork a coin.

6. Security is not gained by skin in the game. Security is gained by distribution - ensuring it is a) harder to hoard hardware and b) ensuring participants are in many areas and locations with different profiles and businesses. GPUs are bloody hard to hoard at scale.

7. FPGAs are arguably harder. ASICS are simple and most altcoins suffer from ASICs belonging to one-two parties (HNS has a huge issue with this). BTC doesn’t have this issue because it has SCALE. Too many participants.

8. BTG was 51% attacked for a few pretty specific reasons: 1. it was a low market cap coin with not enough participants. They didn’t have scale, so it was easy to amass hardware. Even if they had been on ASICs, same thing would have occurred.

9. As long as HW can be rented (and all hardware can), a 51% attack will happen.

10. Then, 2. It had generated a lot of hype and had some pretty senior companies backing it early on in the shadows. Attacking it was “taking it down a peg” and it was also just EASY. It was tempting. Those of us who hung in the right IRC channels at the time-

11. -saw it happen (and the motivations around it) in real time. Attacking it took some of the buzz, invalidated the work of the main team, and crippled them from exchanges. This would have happened even if they were in ASICs. Why? Golly gee, rental attacks are a thing!

12. BTG recovered as they gained more participants. Arguably today they have enough distribution and the space has matured to the point where 51% attacking is not going to generate an attacker any value - where do they sell their ill-gotten gains? OTC? Nah.

13. I could rant about all of this for a long time very angrily because people still think mining behaviour in BTC applies to other coins. Different ecosystems. The miners of one coin behave and act differently from others.

14. Think of a coin like a product. Your customer behaviour and customer study on Ferrari is not going to apply to your Ford pick-up truck. Ok I don’t know cars, so that analogy was probably poor to pick, but you get the gist - different customers, even though they share a label.

15. Perhaps one day, I’ll write a long form article about all of the little attacks I’ve observed and share all the little logs collected along the way. Much of my thinking and statements here have been shaped by real-world exposure to these attacks and the attackers.

16. Arguably, hacking and its psychology isn’t something a researcher can do a cost analysis on - the same applies to mining attacks. Many times, they happen because it’s fun, it’s possible, or it’s out of spite. It’s a nuclear attack, with very little in-person consequences.

17. That’s changed recently with the exchange infrastructure growing, but it is still nascent.

18. To finish a good rant: wtf, CoinTelegraph?