Breaking the Silence on ChatGPT

ChatGPT is a conversational AI system that uses a large-scale neural network model to generate engaging responses to user inputs. It is based on the GPT-3 architecture, which is one of the most advanced language models in today’s world. ChatGPT has been trained on billions of words from various sources including books, news articles, social media posts, as well as web pages.

I have just used it since December of last year, few weeks from its initial release as a research preview and with that being stated, I can testify that ChatGPT is not perfect. It has some limitations that need to be addressed and discussed openly. In this article, I am going to break the silence by sharing some limitations that I personally had encountered and I’ll finally finish with my own views on the ongoing AI race among the big tech companies out there.

ChatGPT sometimes responds inappropriately

One of the main challenges of chatGPT I have just encountered is that, sometimes it could just give you inappropriate and misleading responses. I asked Pavel durov’s educational background and unfortunately it retuned the following response:

ChatGPT resposnse on Pavel Durov’s college degree
source: ChatGPT

While the truth is that, Pavel Durov admitted to St. Petersburg state university for a Bachelor degree in English philology and translation.

Another one is , I asked chatGPT to provide links for video tutorials on YouTube on how to host a flask application and it ended up to provide incorrect links to unavailable videos with wrong titles.

ChatGPT response on tutorial to host flask apllication on GCP

On top of that, I don’t recommend it for research purposes as it generates wrong citations to papers that don’t even exist. Few weeks ago I just wanted to include few case studies to my writing and I asked chatGPT to provide them with the references and of course it did but when you click the link provided then it doesn’t direct to relevant sources!

You better look for AI powered tools for research like Scholarcy, consensus and others which are much reliable and trained to serve you for research purposes.

Generally, despite the fact that over 90% of the responses where accurate but what if you are preparing something serious and ending up with misleading fact(s)?. But the coming of chatGPT has shown great potential and promise for various applications in different domains. However, I hope that GPT-4 has made more improvement and added the precision to the responses it provides.

AI Technology needs a regulations

We haven’t reached the singularity but I hope we are on the way to , the ongoing AI race among big tech companies is my indicator as they put bulk budget and investment to develop the technology and none of them takes some ethical and safety into account as Sam said. Google has just released Bard recently but seems to be in a competitive view to its competitors and I can conclude that Google just released that to answer OpenAI and Microsoft and not otherwise. In fact, these technology will victimize some people’s career and would lead to massive layoffs in workplaces and this should be considered. On top of that, victims need to upgrade their skills and how to utilize the technology in order to achieve a maximum productivity.

Our educational system needs some upgrade

The arrival of AI has just exposed our educational system is out of time as students used the technology to do assignments, writing essay and others. Here not blaming OpenAI for chatGPT but we should now move on to educational system with a way of assessment which measures someone level of understanding and grasping the concepts and how could it be applied daily life. Students should also note that their future competitor isn’t a human being but the AI and by knowing that they should invest much more in deep thinking and creativity than traditional memorization way of learning.

Final words

it’s not bad to develop powerful AI technology but we should remember that, the technology concerns the humanity in sense of enhancing the human potential. Now we need more powerful AI to monitor environmental pollution, climate change and other related affairs.

Leave a Comment