This Op-Ed was written by Marco Bianchini, Economist of the Centre for Entrepreneurship, SMEs, Regions and Cities of the OECD and the Coordinator of the “Digital for SMEs” Global Initiative.

Please note that this article has been originally published here on the OECD “COGITO – Insights on Entrepreneurship, SMEs, Tourism, Regions and Cities” portal and is presented here with the authorisation of its author.

ChatGPT and other large AI-based “generative language models” are taking the world – and our news feeds – by storm. As Beta versions of the software are made available to the general public, observers are getting excited about the capabilities and the vast array of potential applications. But what does this mean for small businesses?

The impact of ChatGPT on small businesses could be significant, depending on how the technology is used. ChatGPT is a large language model that can generate human-like text, and small businesses could use it in a variety of ways, such as:

1. Automating customer service: ChatGPT could be used to train chatbot and virtual assistants for various tasks such as answering customer queries, providing product recommendations and helping with purchase transactions, which could help small businesses save time and money.
2. Generating marketing content: ChatGPT could be used to create engaging, high-quality content for social media, e-mail marketing, and other channels, which could help small businesses increase their reach and engagement.
3. Improving product descriptions: ChatGPT could be used to create detailed and accurate product descriptions, which could help small businesses improve their e-commerce sales.
4. Language Translation: ChatGPT can assist small businesses to translate their content, product descriptions, and other materials into different languages to expand their reach to new markets and customers.
5. Content Personalization: ChatGPT can be used to generate highly personalized content for individual customers and prospects, which can help to increase engagement and conversion rates.

Overall, ChatGPT has the potential to help small businesses increase efficiency, reduce costs, and improve their marketing and customer services efforts.

Whatever you think about the paragraph you just read, read it again. Because it has been 100% generated by ChatGPT in just seconds. It answered my question, “what will be the impact of ChatGPT on Small Businesses?”. Indeed, it’s threatening to put this blogger out of business.

0 to 100 million 

ChatGPT has become the fastest growing internet application in history, reaching 100 million users in just 2 months. It took 9 months for TikTok in 2016 and 1 year for Google in 1999. Its language model (GPT stands for “Generative Pre-trained Transformer”) uses 175 billion parameters to make predictions and generate “human-like” text based on available information on the internet up to the end of 2021 (~45 terabytes of text data).

New uses are emerging every day, with international surveys suggesting that at least 30% of white collar professionals have tried the tool at work and MBA professors confirming the tool was able to pass one of its exams. Speculation is also emerging on its potential impacts on internet search engines and how it might be used to hijack democracy.

For entrepreneurs, the question is if and how they could integrate “generative language models” in their operations. The examples above relate to functions such as “customer service”, “marketing”, “translation”, activities in which efficiency improvements seem plausible and immediate.

SMEs will need skilled personnel to integrate this tool effectively, and to understand its limits as well as its potential. This is a common challenge for SMEs, which together with the lack of financing and of management skills to implement change, results in a consistent gap in the use of advanced data analytics/AI software between SMEs and large firms (OECD, 2021).

Small firms lag behind in the use of big data

Businesses having performed big data analysis

Source: The Digital Transformation of SMEs (OECD)

Vanishing vocations?

Even where successfully deployed within firms, there are broader risks to manage for policy makers, including its potential impacts on the world of work. There are familiar fears about of “disappearing jobs”.

Past OECD research suggests it may be more appropriate to think about “disappearing tasks within jobs” than disappearing jobs per se, with even jobs at “the highest risk of automation” still having only 18 to 27% of skills and tasks that are amenable to automation. This is likely to be true of Chat GPT too which could help white-collar workers automate boring tasks, freeing up time and resources to be invested in more cognitively intensive activities.

Yet while this move up the cognitive chain may be welcomed by some workers, evidence suggests that the increased mental load can also increase work-related stress in companies adding to a mushrooming mental health crisis. This involves rolling out the right support to tackle rising levels of stress in the workplace, building on the fine example set by New Zealand.

ChatGPT may widen income inequalities as workers with higher digital and cognitive skills harness the power of AI to move up the value chain while other workers are left picking up the pieces of jobs left behind by the bots.

So what can governments do? The software itself couldn’t point at specific policies when asked directly, so – for now – this blogger was relieved to find there is still a role for OECD analysts… and national policy makers.

Working with countries through the OECD Digital for SMEs Global Initiative, we are encouraging governments to adapt SME business advice and support programmes to embrace the potential of AI solutions and to help them manage the risks.

Read more on the OECD work on Digitalisation on SMEs.