The Mexican Blue Oyster, or if FOMO should be a driver for disruptive AI

With ChatGPT release back in November 2022, a revolution might have started in the way how persons and business obtain information, make decisions and plan or perform work. Despite the technology still has issues, it has already shown amazing results, like passing graduate law and business exams, solve code problems, or even make biblical verses from nonsense prompts.

Some Universities are now taking precaution measures to ensure that student assignments are not performed by AI, and as a trick someone apparently made ChatGPT to write this. Funny or scary? Your call!

This has caused such an impact that many are predicting the end of search engines, which of course caused a quick reaction from the tech giants Google and Microsoft to launch their own ChatGPT based AI, however the introduction to their new flagship AI powered search bots hasn’t been the best so far.

The Mexican Blue Oyster is a reference to the AI powered Bing search demo, where the AI responding to a Mexico city nightlife planning, fabricated some details about existing bars and clubs and missed important facts on others, like one of the recommended places is the oldest gay bar in Mexico! While it might not be a harmful suggestion, I can’t help to associate this with the famous Police Academy movie scene, where things get messy after the cadets tricked their Sargent to meet at a “salad bar” which after all was quite a different place. Still funny after all those years!

Should FOMO be a driver for disruptive AI

Funny stuff aside there’s some serious issues at hand here, such as incorrect financial reports, fabricated information and even some reports on erratic and concerning behavior. There are several online articles I’ll just leave this one as a resume.

As OpenAI Chief Executive Sam Altman tweeted this is still a technology in progress, however it’s achievements have caused the tech giants to race, with obvious concerns that their search business might just have went through the bin. And what this reaction have caused? A rushed product based on FOMO, or “Fear Of Missing Out”, that shows something, but is still a long way from a final product.

The main issue to understand here is what exactly it can produce. ChatGPT doesn’t exactly know anything. It’s an AI trained to recognize patterns in vast sources of information gathered from the internet, with additional human assistance training through a model using Reinforcement Learning from Human Feedback (RLHF). While the answers you get may sound plausible, they might well be entirely wrong, since this AI chat model is expected to complement information based on the gathered data without knowing if it’s true, exact or missing key facts.

When the tech giants (apparently?) knowing it’s limitations, start to advertise and perform demos like crunching corporate financial numbers, or trip plannings, they’re selling a concept that this iteration of AI is not the best at, and can cause unexpected results. This really gives me the vibes of “Let’s go agile” fever that took over some years ago, or “let’s microservice everything” because everyone else is doing it, or “insert xyz tech fever” that you’ve adopted without exactly knowing how to do it or if it’s actually a good path.

Why this happens? FOMO, or Fear Of Missing Out.

And while the band is playing and everyone’s really excited about the great innovations ahead of us, someone must ask the really important question:

Such potentially disruptive technology, made from tech giants with a global reach, that can provoke an enormous impact on people, business, governments and entities, should be driven from FOMO and rushed into the public ?

What if you make a political question and the answer is biased? What if it’s about race, religion, equality or freedom of speech? With the potential to influence the work, opinion and daily life of persons and companies across the globe, what fail-safes are being implemented to ensure that this technology cannot be manipulated or distorted? This is why a disruptive technology like this cannot be rushed.

Don’t get me wrong! I’ve been a fan of most of Microsoft and Google products. In their own times, Windows, Gmail, Google Maps or Microsoft Teams, just using a couple of examples, have revolutionized the way we live, work and interact with each others and the world. The innovations they have delivered are amazing, however in this one I’m not going with the fan-base claiming that our live has just changed.

AI is here to stay and already has great applications. Sollogica is a partner of Hubspot, which uses AI to improve user experience, however despite the recent advancements, AI chatbots still have some limitations and people should be advised to use them with caution instead of jumping straight on the bandwagon without understanding it’s dangers and limitations.

Who knows, maybe one day you’ll end on your own version of the Blue Oyster!

Leave a Reply