What if AI technology joined forces with the generative chatbot ChatGPT?
It is no secret that the world held its breath at ChatGPT's latest launch. Neither is the fact that while some eagerly use the application (and its siblings such as Bing Chat and Google Bard), others are sceptical. As CEO of AI software company Nalantis, I can’t tell you how many times I've been asked about it. Remarkably, most questions have a negative connotation. Whether we see it as a threat, whether I myself think AI will make jobs disappear, whether I’m aware of the dangers of chatbots and generative AI?
To jump straight to my answer: generative AI offers benefits for everyone. From creatives, developers, legal officers, administrative staff ... to even the team at Nalantis.
NEC PLUS ULTRA
After more than fifteen years in the technology industry, I’m convinced that ChatGPT, Bing Chat and Google Bard are strong and high-performance data processing tools. They are no doubt the nec plus ultra in generating smooth data-based language.
But within that strength lies an important nuance: ChatGPT’s is a generative technology. A collection of statistics and word sequences that allows it to write coherent text, following your input or question prompt. However, the software is not trained to semantically analyse the data it finds. Let alone verify its relevance and trustworthiness. In other words: it does not understand what it generates.
That is where ChatGPT most differs from Nalantis: we make analytical technology. We are at the root of the path that ChatGPT should be following. Our technology provides in-depth analysis so that tools like ChatGPT can generate more accurate and relevant content. What’s more, large language models like ChatGPT act statistically, but Nalantis works semantically or based on meaning.
Not a threat, but an opportunity
ChatGPT cannot replace Nalantis' technology, but it can support it. Two examples:
Thanks to our application for cities and municipalities, officials and citizens find answers to a specific question much faster. This is because our Natural Language Understanding technology analyses the relevant administrative documents based on their question. Currently, the application only displays the found documents and the questioner has to filter his answers from them. If we were to add ChatGPT as an additional layer to this, the chatbot can generate a well-written answer from the documents found by Nalantis. We can then be sure that the answer contains the correct and relevant info, as it has been found by our analytical AI. However, this is not possible at the moment because ChatGPT cannot yet handle this small data application.
In our Talent Acquisition Platform, we analyse candidate-related documents inputted during the application process. Resumes, video resumes, test results, interview reports ... Nalantis could pass on its analysis of this to ChatGPT, which then writes a clear report on the candidate's profile. Same but different: Nalantis provides the data, ChatGPT processes it. Unfortunately, this also is not yet possible, but something to look forward to.
Besides figuring out future integrations with other AI applications, our R&D team is exploring how ChatGPT can help us write code faster and semi-automatically create new linguistic resources. Because yes, ChatGPT does more than answer your questions. With cleverly crafted prompts, you can make the chatbot perform tasks that make your workflow a lot smoother.
There’s no denying it: as an AI technology company, we don't see ChatGPT as a threat, but rather as a potential support of our services. Because although you can’t put Nalantis and ChatGPT on the same level, they might help each other climb higher.
Written by Frank Aernout, CEO of Nalantis
Would you like to read more and stay updated on new developments?
Connect with Frank on LinkedIn.
