4/5 (1)

Last week, the cryptoworld was thrown into chaos as a report from Hategan (2023) showed that the vast majority of NFTs had become almost completely useless. In the report, Hategan (2023) mentions that currently 95% of NFTs have a market cap of 0 ether, and thus are worthless, while only less than 1% of NFTs is worth more than 6000$. This downturn in the market was mainly caused by the increasing supply of NFTs while the demand remained to be the same (Hategan, 2023). It eventually resulted in a buyers’ market where NFTs are more critically assessed by investors (Hategan, 2023). 

The increase in supply can partly be explained by the development of visual generative Ai tools like ChatGPTs’ DALL-E 2 which is able to create realistic images based on a text description (OpenAi, n.d.-a). This meant that creating an artwork to sell as NFT became easier and more accessible for the majority of people. Take, for example, the image down below which was created in DALL-E 2 and could be easily converted to a NFT and sold on the market.

Figure 1 | OpenAI (n.d.-a)

This evolution of visual generative Ai tools will continue to develop as OpenAi will introduce their new version of DALL-E in the near future (OpenAI, n.d.-b). DALL-E 3 is said to need less text prompts than its previous version and also be able to generate images of far better quality (OpenAI, n.d.-b). It will therefore be able to create artworks that are even better and easier to generate than currently possible which will drastically increase the supply and quality of NFTs. 

Figure 2 | OpenAI (n.d.-b)

Eventually, I think this will transform the NFT market from a highly volatile and speculative market to a market where NFTs are valued on the genuine artistic value and the utility it holds. Furthermore, I believe that it will also boost the exploration of other applications of NFT technology like using NFTs to verify certain certificates and academic degrees instead of being used mostly for artwork. 

References

Hategan, V. (2023, August 29). 95% of NFTs are Dead – Trends, Predictions & Statistics 2023. dappGambl.com. https://dappgambl.com/nfts/dead-nfts/

OpenAI. (n.d.-a). DALL·E 2. Retrieved September 26, 2023, from https://openai.com/dall-e-2

OpenAI. (n.d.-b). DALL·E 3. Retrieved September 26, 2023, from https://openai.com/dall-e-3

No ratings yet.

With the world discovering more and more about generative Ai, we start to come closer to understanding the impact it will have on our future. Lately, I had also been discovering more about generative Ai myself when I started using ChatGPT to assist me in multiple activities due to its wide array of functionalities. One of these activities was creating an appropriate training scheme for running the marathon of Rotterdam in April when I discovered that the output of ChatGPT greatly differed based on what I asked it to generate. For example, when I just asked ChatGPT to “Create a training scheme for a marathon” I got a very basic output but when I started giving it more context based on my height, weight, heart rate zones and lung capacity the output was far more tailored to my specifics than before. This is also known as prompt engineering; the process of creating specifically tailored instructions to achieve the best possible output (Cambridge Dictionary, 2023).

But how valuable is this skill exactly? Well, that is still pretty hard to determine as it has only been around for such a short time. Short and Short (2023) for example, really emphasize the importance of learning to provide generative Ai models with well-designed instructions to utilize these models more effectively. In their research, they show this by creating and refining different types of pitches based on CEO archetypes which has shown great results. However, Acar (2023) tends to think differently of the importance of prompt engineering. He, for example, states that problem formulation is a far more crucial skill as it isn’t bound by the same constraints as prompt engineering which include its dependence on the model used, its requirement of proper linguistic skills and its shortcoming of adaptability across different problems. Furthermore, he also argues that Ai models will evolve in their natural language skills and thus require less tailored instructions to provide the best possible output.

So what do you think? Is it worth investing our time in? Or should wait and see how generative Ai will transform over time before jumping fully in improving our prompt engineering skills?

References

Acar, O. A. (2023, 8 juni). AI prompt engineering isn’t the future. Harvard Business Review. https://hbr.org/2023/06/ai-prompt-engineering-isnt-the-future

Cambridge Dictionary. (2023). Prompt Engineering. https://dictionary.cambridge.org/dictionary/english/prompt-engineering

Short, C. E., & Short, J. C. (2023). The Artificially Intelligent Entrepreneur: ChatGPT, Prompt Engineering, and Entrepreneurial Rhetoric Creation. Journal of Business Venturing Insights, 19, e00388. https://doi.org/10.1016/j.jbvi.2023.e00388