Strona korzysta z plików cookie w celu realizacji usług zgodnie z Polityką Prywatności. Możesz określić warunki przechowywania lub dostępu mechanizmu cookie w Twojej przeglądarce.
Each of us searches for many things every day. Often we don't realize how well search engines have come to understand our queries - thanks largely to AI. Google, for example, has improved significantly over the years, making search more intuitive and natural. Most notably, in 2019 they introduced BERT models. Leaving aside the technical details, it’s an intelligent language processing system. It's by no means sentient AI, but when it comes to searching, it understands what your questions mean, producing more relevant results than ever.
Machine learning algorithms also help filter spam and duplicate results, as well as differentiate between good and low-quality content, limiting the latter.
The e-commerce market has grown significantly over the past decade. People are buying more products online than ever before. AI is contributing to this growth - helping companies analyze the vast amounts of data collected from online consumer behavior to provide predictive insights into sales, which are used in part to make sure warehouses are well-stocked. They analyze demand and improve supply chains, making the entire process more efficient than it could ever have been without intelligent systems. For the end user, this results in much less frequent shortages, quick delivery and lower prices.
It's really convenient to order a ride, knowing exactly when it will arrive and how much it will cost. However, this has not always been the case. Ride sharing companies such as Uber, Bolt, Lyft ect. are investing heavily in AI technologies, optimizing their services on many fronts. AI allows them to forecast supply and demand during busy periods, adjust rates accordingly, and more precisely estimate pick-up and arrival times. Uber also uses historical data to more accurately match passengers with riders.
Let's say you are watching an interesting video about modern aircraft engineering, imagine going down the youtube rabbit hole. You hear that the PW1000G engine has achieved the highest bypass ratio in the history of commercial aircraft. I know what you're thinking. What on earth is this "bypass ratio" and why does it even matter? The suggested videos will most likely explain it to you. If they don't, you start typing in your query and before you know it, you're linked to the right film.
Recommendation systems have been around for a long time, but recently they have become much smarter, thanks of course to the vast amount of data on user activity. Companies can match a user's tastes and interests much more accurately than before the era of deep learning. This has a lot of positives, for example, on Netflix you are very likely to be recommended movies that you would actually enjoy.
Unfortunately, there are also drawbacks. Apps like TikTok do incredibly well at selecting short videos that grab your attention. This can keep you glued to your phone for much longer than expected, providing often not very valuable content. Systems like this are the reason you end up watching videos about modern aircraft and aircraft engineering in the first place, even if you're not really concerned with it.
In recent years, the use of AI has revolutionized smartphone cameras. Instead of the focus being solely on bigger and better hardware, software has become a very important component. Intelligent photo processing is detrimental to achieving good results. I'm not talking about fancy Snapchat-style filters, but more about the overall improvement in photo quality that AI quietly performs every time you take a shot.