Things AI is good at and things AI is bad at

There is no denying that AI has changed our lives. But as with anything new, there are positives and negatives

AI HAS GONE from a pretty obscure and theoretical concept to something that is everywhere in the space of a few years. Like with anything new, there are positives and negatives to artificial intelligence. Here are some examples of where it is a help, and where it is a hindrance.

One pretty niche, but also pretty cool, area where artificial intelligence can help people out is in the world of gambling. For example, an AI program can help a player find games that suit their preferences, playstyle, and tolerance for risk. Since some crypto casinos have a lot of games, whether it’s card games, dice, poker, blackjack, or any number of virtual slot machines, this saves time and frustration and also allows players to focus on playing the games they’ll want to play.

But it isn’t just game selection that makes AI a useful tool for gamblers. If you are playing at one of the best online crypto casinos in Canada, then the chances are that AI is at work, monitoring the blockchain to make sure that transactions are fast, secure, legitimate, and accurate. And it can do that much faster than a human analyst.

For instance, something that has earned crypto casinos a good reputation among gamblers is that they offer instant withdrawals. For instance, if you’re used to conventional, traditional, online casinos that can make you wait up to a week before you have access to your money, winning a game and being able to take your money from your account straight to your crypto wallet can seem like a breath of fresh air. That is all thanks to AI. AI can analyse the transaction instantly and verify confirmations, fee settings, and wallet validity in seconds, which is good for gamblers. Similarly, AI can find out if there are insufficient gas fees for the account, or if someone is trying to use a crypto that isn’t accepted by the casino. This means that it can, in effect, fix problems before they even happen, which makes for fewer unconfirmed or delayed transactions which, in turn, makes for a better, smoother, and more enjoyable gambling experience.

Another place where AI can be useful is transport. In some ways, people have been using a kind of AI when it comes to GPS devices. If you’ve ever been on a route but have been diverted because your SatNav has found a quicker way to get somewhere, that’s pretty much AI. Likewise, some companies, such as Amazon, are using AI to help automate their deliveries, meaning they can be safer, quicker, and cheaper.

One area where AI is more of a hindrance than a help is writing. While using AI can, at first glance, be helpful, it can do more harm than good. The problem with a lot of large language models (LLMs) is that they only really have one sort of voice. This often means that people who use AI to finalise their work end up sounding the same as everyone else who uses AI to finalise their work. While style guides are common place in publications, this takes uniformity of style and turns it into uniformity of voice. And, since the likes of ChatGPT are based on the American form of English, differences in spelling can catch people out and make it look painfully obvious that someone hasn’t actually bothered to write something themselves. It becomes even more obvious when the person can’t properly write in the language they’ve asked AI to generate something in, too. That’s not to say that AI isn’t a useful research tool, but it can’t replicate a properly written article, letter, or post.

While using an LLM for academic or personal writing is bad, it has infected the internet to such an extent that, according to Wired, more than half long-form English language posts on LinkedIn are AI-generated. And when you’re in a place where everything reads the same, eventually it will all merge into one amorphous mess. This is bad for a number of reasons. Firstly, good, important, things get lost in the shuffle. Secondly, to put it bluntly, if someone can’t be bothered to write something, then why should anyone bother reading it? And there is a very good argument that people who use AI for simple Slack and WhatsApp messages don’t deserve to have the internet.

But that’s enough about that. Another place where AI is a hindrance is in the job market. While having AI to allow candidates to apply to jobs quicker and having it whittle through applications sounds great, it brings a whole new stack of problems. For example, it is now much easier for completely unqualified candidates to apply for jobs. This means that jobs that might have once had 10 or so applicants will now have 100. And, with the work of HR people and recruiters rising as a result, they will be forced to use AI to find suitable candidates. But, because there isn’t any human oversight, not only will companies end up overlooking perfectly qualified applicants whose applications aren’t perfectly optimised, they will also end up finding that they can’t hire anyone suitable because their HR departments haven’t actually taken the time to read through applications. So, if you’re struggling to find work in 2025, AI is a definite hindrance.