With each passing year the amount of data, both structured and unstructured, is increasing exponentially. When it comes to polling, there was very little data in the early 1990s and 2000s and opinion polls were usually way off the mark. That changed a bit in the 2010s when predictions started getting more accurate. However, the 2024 general elections proved to be error prone when it came to opinion and exit polling. Most agencies pointed to the BJP gaining seats when it in fact it lost a substantial amount of them.
Despite the fact that there are umpteen polling agencies, hundreds of millions of citizens giving their opinions day in and day out and a social media which is overloaded with myriad opinions all the time. It appears that too much data is as worse as too little data for that’s when everyone cherry picks what they want and prediction becomes eventually biased.
The US Presidential elections appear to be much worse. Every week you may find a headline which is totally contrasting on who is going to win. In one news aggregator site while one headline said Donald Trump was heading for a landslide, the very next suggested that Joe Biden was sure to win easily.
In fact, in many a case you will find that the predictions are 50-50, basically a coin toss. After the results are in, the wrong predictions are hidden, and people forget them. The right ones are touted and magnified, and we move on to the next cycle repeating the same mistakes.
Clearly human beings are totally inept in handling a huge amount of data. Sifting through it and identifying the right ones and analyzing them is a herculean task. With a population decline and digital data overload, is AI ultimately destined to sift through and anaylze all data? That is a scary thought because with each passing year we will be more and more at the mercy of AI. That’s why many science fiction writers have predicted that an AI entity will ultimately be the ruler of the world.
ChatGPT has given a glimpse of how much we like to be dependent on AI to make our lives easier. Home assistants are slowly taking over. Tesla has come out with a humanoid robot called Optimus (or the Tesla bot). The Gen 2 version was showcased recently, and one realizes how a “Robot+GenAI” could be a game changer when AI could just take over every home in the world. While the $10,000 price tag being estimated seems expensive, that would make it cheaper than a car and in the long run the costs could come down.
At this stage AI just seems inevitable.
The inevitability of AI
It appears that too much data is as worse as too little data for that’s when everyone cherry picks what they want and prediction becomes eventually biased.
New Update