I came across a story of a man who had lost access to a wallet containing about five Bitcoin, which he bought at $250 each years back when he was in college. For several years, the coins had been inaccessible, because he had changed the wallet password when he was drunk and couldn't remember it.
This thing happened in 2015, and since then, began a long, frustrating ordeal. The man tried every solution he could find to regain access to the coins, running brute-force attempts on his computer for years, testing every conceivable combination of passwords and even employing experts to crack the wallet, but his wallet remained locked despite millions of possible password combinations having been tested.
He eventually had to accept the loss because there seemed to be no apparent solution. The Bitcoins remained locked in his wallet, since there's nothing more he could do about it. As a one more effort of trying his luck, he tried something new, he decided to use Anthropic's Claude AI and fed the computer files from his old college laptop, in the hope of finding a relevant clue. The AI then found an older wallet file, located a flaw in thebtcrecovertool that he had been using all this time and enabled him to crack the wallet, recovering years of lost Bitcoin.
Those five Bitcoins are now worth nearly $400,000 and the most astonishing thing about this story was not that the man got his money back, but how it happened. It was not a case of trying harder, but a case of AI being able to find the correlation between lost pieces of data, identify problems that humans can't and ultimately find a solution that has remained locked away for several years.
Before now, I have been been very cautious in the area of AI usage, and have always felt that we should be very careful about it usage especially with areas related to human life, but, after seeing something like this, I am beginning to have a change of mindset. This goes beyond simply saving time, or automating tasks; this is about capabilities on a completely different level of the problem solving.
If AI is able to take years of stored data, find a common factor within it, discover hidden issues within something and produce results like this then I feel it is more than safe to assume that we are entering an age where artificial intelligence can achieve far more than what many of us currently believe. This story has proven that AI are not just for performing simple tasks or for automation, but are a tool which we can use for the most complex tasks that are beyond human capacity.
From now on, I support the AI evolution, and although I say this not blindly, I feel that it is clear from this particular experience that a paradigm has already shifted. The real question now is not whether we use AI, but how best should we do it, because there is only going to be more of it and we should probably be quicker to adopt the technology.