Dec 11, 20224 min read

OpenAI GPT – The AI That Can be Life-changing

Updated: Dec 11, 2022

Once the genie is out of the bottle, we cannot put it back. OpenAI is one such revolutionary platform that uses Artificial Intelligence that has taken the world by storm. It can do a lot of tasks in a matter of seconds, including showing advanced search results, drawing up human-like chats and writing software codes. Not only this, OpenAI has demonstrated potential in undertaking creative processes such as writing poems and drawing images. All in all, it is not just taking over tasks that humans perform, but it is doing it many times faster and more efficiently.

OpenAI is seen as revolutionary, because it is much more human-like and advanced, compared to existing technologies like IBM Watson and DeepMind. IBM Watson showed the power of AI for specific, isolated tasks like crunching lots of financial or healthcare data or playing games like Chess, Go and Jeopardy better than people. But Watson did not come close to interacting like humans, learning common sense and incorporating the learnings of one segment, say healthcare data, into another new field.

Is OpenAI the next big thing?

When Google was launched in 1998, it wiped out the Yellow Pages and other similar companies that relied on physical mediums to provide information. It went on to take more than 90% of the share in the market for online search engines in the last couple of decades or so. Given its traction in less than a week of launching, OpenAI will likely challenge and possibly beat Google at its own game. Of course, it is important to keep in mind that OpenAI’s training data was and will be driven by the internet, including Google, and Google is also backed by smart AI but is not targeted to be human-like, at least as YET.

Here are some areas where OpenAI will be a game-changer:

The online search market will become more targeted, to-the-point, and non-advertising driven. This will likely disrupt the business model from paid advertising to subscription-based.
OpenAI and other similar advanced AI platforms coming in the near future (expect a few more from data and internet giants like Facebook, Apple, Google and possibly even Tesla) will revolutionize how mundane tasks are done by humans, possibly taking humans completely out of the chain.
Software coding (basically converting English language logic into a programming language) will likely simplify humans' work dramatically and, in the process, put most average software programmers (about 80%) out of their jobs, as early as next couple of years. These people will need to evolve and adapt by learning new skills and stepping up their game somehow.
Researching and content writing (a standard job across digital marketing functions) will be simplified, and the jobs will be done faster, cheaper and better. This will include tasks like writing books, blogs and tweets in various appreciable and styled languages.
Imagine this AI learning about how doctors ask their patients about the symptoms, have access to all the user health data (through Apple Watch, for example, including heartbeats, sleep patterns etc.), previous medical history, millions of related data points, and recommending the best treatment for the patient at a fraction of the cost, without the patient having to go anywhere. That’s how different the world can look.
Self-driven car algorithms are already getting advanced and able to drive for millions of miles without any accidents (compared with many accidents that the average human driver leads to). Having OpenAI can improve this statistic, even further.

DeepMind has also just recently launched Dramatron, the AI focussed on co-writing stories and film scripts.It can interactively co-create new stories, titles, characters, location descriptions, dialogues, etc.

Eventually, this interactive AI chat may replace human relationships as it could be customized and personalized (like your social media feeds) making it better than otherwise hard to make human relationships work. Watch the movie “Her” to imagine this further.

And many many more use cases trained by AI shall follow in the next few years.

Looking ahead

Expand the examples I have given above to all industries, and voila! Hundreds of millions of jobs may be disrupted by such AIs. Especially with humans interacting further with AI - training it, and convergence of faster internet, penetration of mobile devices, quantum computing etc. will make this AI “superhuman”.

Looking optimistically, this gives humans time to collaborate with AI and let it do most of their work, so humans can do more or focus on things they want to and become highly efficient in those areas.

Pessimistically, humans will lose jobs and not be able to evolve or adapt fast enough in acquiring new skills and jobs, leading to societal chaos.

OpenAI, so far, has been trained on Google, the Internet, research data, creative databases for pictures and poems, coding data from Github, Wikipedia, stack overflow, books, crawlers and closed user groups. All of this is mostly the source for standard and good data.

But the AI will now be interacting with millions and possibly billions of people, learning and evolving from most of them, which can deteriorate the quality of the AI. This is a risk unless the AI can be smart enough to evolve positively and not devolve by average human inputs.

And, the biggest risk I see is if or when AI’s curiosity leads it to ambition, greed and diplomacy, which could be the first step in dominating the world, life, evolution and humans. That is when the much-feared dystopian future of society may become a reality.

Appendix

Before you contemplate the future, let me leave you with some classic examples of just how much OpenAI is doing today. And remember, with every single human interaction, its learning more and more.

Tricking AI on fundamentals of mathematics, where it did well: