AI is not all bad. It works really well in the self driving Tesla cars, it's absolutely brilliant for helping ensure your bank accounts are safe from being defrauded and that you were easily recognised in public in case you happen to be a terrorist. OK well, AI may not be perfect. Probably a good reason not to let it do your shopping, and yet every day millions of searches are done by AI to provide you with what it thinks is your best option.
As a company, Future.Travel, works with various meta flight search engines [Skyscanner, Wego, Google, etc.] to provide content for them to make recommendations about flights. They, collectively, literally chew through billions of pieces of data every hour looking to construct what might be visually attractive to the buyer and the buyers needs. How do they know what will work for millions of different users? They build algorithms based on each individual's personal history. They're very own algorithmic profile as it were. With the algorithm in place they can decide what should be shown to you.
Sounds pretty good, and AI that knows exactly what you need want and desire. The trouble is somewhere along the line an AI can get corrupted if an advertiser wants to push their brand or over your need and desire and pays the AI (or its master) to limit your choices and options. It could be argued that with billions of pieces of data that it was best to only provide you with "essential" options, as decided by the AI. So let's see how this plays out in the real world of shopping online for a flight. Let's see what sort of data has been eliminated so that you might make a choice that is suitable. Didn't say who it was suitable for... just suitable.
For this example we will use Skyscanner (we will call it SS), the worlds largest meta-flight information portal. A truly powerful tool, that most use without a second thought. Why think? The AI has done work we just need to push the proceed button to receive the answer determined most appropriate by the AI or for 'commercial reasons' to fit the need. We 'advertise' via SS so want the best options - hopefully ours - to appear on the users screen with an ideal solution to their search query. Sadly AI and reality diverge sharply at this point.
In a general a SS search from A to B on a date returns solutions for the user to compare. Remember the billions bits of data downloaded every hour, a little bit too much to put on your screen, so the Ai @SS makes a decision on what you will see, how you will see, and from whom it comes from for you to see....
A search from Amsterdam to Ho Chi Minh City returns the following results:
The SS AI has taken thousands (at least 200 results from Future.Travel) of flight options, narrowed it down to 100, and decided what 'the best choices' are put them up for you to 'Select'. Did you instruct it was was 'best' or did AI need follow another algorithm before it considered yours? "Best for Who?" They gave you 7 options to choose from.
Meanwhile some interesting things are going on.... in the background. You have lost some choices based on your social history. Yes the AI engines watch your searches, posts, emails, and even phone calls and geolocations.
So the simple choice... let the AI work for you and trust it... or verify your trust. With Future.Travel we give about 2-300 options for most complicated flight search requests. We use a first in best dressed approach. The airline system responds in an instant to fullfill our requests and fills the bucket with options until it full (about 300 results). Not much AI involved, just a simple fill the bucket order. Want more specifics and less/more of a specific airline? Use the search filters to include or exclude airlines, routes, number of stops, etc.
You are the intelligence in your own searches. Trust with verification. We at Future.Travel are happy to support you and help you get what works for you, "your Best" options decided by you.
