Fundhouse Insights Latest Articles

What language do you dream in?

Published: March 30, 2023 by Fundhouse

As technology continues to advance at a breakneck pace, we are witnessing the emergence of Artificial Intelligence (AI) in various industries. While this may seem like a positive step towards a more efficient and innovative future, there is growing concern about the extent to which AI will replace human labour. It seems like every day there is a new headline warning of the imminent threat of AI taking over jobs previously performed by humans. It begs the question: are we setting ourselves up for failure as a species? Let’s dive deeper into this topic and explore the potential consequences of a world increasingly dominated by machines. Ironic isn’t it that this introduction above was written by ChatGPT – the poster child for Artificial Intelligence software in 2023. All it took was the simple prompt below and around 10 seconds:

“Write an intro to an article I need to write which covers the idea that AI is slowly taking over some of the tasks that humans currently perform, and that there is broad concern that we are setting ourselves up for failure as the human race. limit the answer to 80 words and write in a style which is formal”.

In November 2022, OpenAI’s CEO Sam Altman announced the launch of ChatGPT. This is some seven years after the company started in 2015 as a not-for-profit, aiming to develop artificial intelligence systems in a responsible and beneficial way. ChatGPT is an intelligent chatbot that helps automate chat tasks. The GPT in ChatGPT stands for “Generative Pre-training Transformer”. The Generative component represents a model that can generate text. Pre-training relates to the fact that the model has an enormous amount of data in it. The transformer component refers to the architecture of the AI model. Hence, ChatGPT simply means that this AI is capable of handling both long and short requests.

Within just five days of its launch, ChatGPT witnessed an explosive surge in users, breaking records with over 1 million engagements. To put this into perspective, it took Facebook 10 months to achieve a similar feat.  This level of take-up was not part of the plan. In fact, Altman and the broader OpenAI team had almost put ChatGPT out to pasture when they decided to release it to the general public to see if real human engagement could help it learn in new ways, what they call something of a “Hail Mary” decision.

This launch has ignited a new landgrab within the technology world, one that has been operating in the background for some time, but which has also been dormant as a productivity tool for the average person.  Generative software such as this seems likely to form the basis of a new technology product cycle in the coming years, as we have seen with other disruptive technologies such as the smartphone and cloud based computing.

ChatGPT has the remarkable ability to engage in lengthy and seamless conversations, provide responses to queries, and create a wide range of written content based on individual requests, such as advertising campaigns, computer code, and humour. Although there have been chatbots in the past, none can match the impressive capabilities of ChatGPT. Here is another intriguing instance that demonstrates the diverse outputs generated by ChatGPT and the level of precision that can be achieved through its parameters.

ChatGPT’s output is truly impressive, delivering the desired results in the expected format and sticking to the specified word count.

The evolution

The journey of technological progress throughout history is marked by significant milestones that have had a profound impact on society. It started with the harnessing of electricity, which paved the way for the invention of the telegraph. This innovation revolutionized communication, leading to the development of the first telephone and ultimately the mobile phone, which transformed the way people connect with each other.

Similarly, the invention of the first personal computer changed the way we process and store information. This breakthrough led to further technological advancements, enabling us to perform complex calculations, automate processes, and solve problems that were previously inconceivable.

These technological advancements have had a remarkable impact on human productivity and efficiency, making it easier to communicate and achieve tasks in shorter periods of time. It’s hard to imagine how the world would function without these innovations today. With each new breakthrough, we see an evolution in the way humans operate and accomplish more with less effort.

But with ChatGPT and the advancement of artificial intelligence in general, the fight is on to see who will own the concept of rational thought – up until now the exclusive domain of the human race.

Displacement of human edge?

As an information source, ChatGPT depends on the historical data it has been trained on. During its training process, ChatGPT analyses massive amounts of data from sources on the open web to identify patterns and connections between words, phrases, and concepts. Through this procedure, ChatGPT can anticipate the probable next word or phrase in a conversation and produce fitting and coherent responses in real-time. As time passes, the model continually enhances its responses through interactions with more users and feedback on its responses. This process is called machine learning. By analyzing the data it receives, the model is able to adjust its algorithms and make more accurate predictions or generate better and more intuitive responses.

While ChatGPT’s design draws inspiration from the neural networks of the human brain, it operates differently from humans in terms of its “thinking” processes. It operates by relying on statistical patterns in the data it has been trained on, whereas humans possess the ability to reason, understand context, and apply logic to make decisions. Furthermore, ChatGPT doesn’t experience emotions, hold beliefs or desires, or make value judgments, unlike humans who are capable of these complex cognitive processes. As a result, the responses generated by ChatGPT are likely to lack depth and insight, making it less capable of providing the nuanced perspectives and insights that humans can offer. Engineers are working on extending the AI technology into this area, so it seems to be not so much a question of if, but when, AI software will be able to replicate most of the more subjective thought processes.

Scientists and university professors have already challenged ChatGPT to several exams across fields as diverse as law and medicine. While it was not the top student in the class, it did manage to pass exams at both Wharton and the University of Minnesota.

Technology has crowded out the inefficiency of human labour for some time, giving us expanded horizons and productivity in return. Yet with each innovation, the hurdle gets raised higher for humans to remain part of the global value chain.

The casualties and the opportunities

AI-based innovation is both a blessing and a curse.  For business owners, increasing your adoption of AI software can improve service levels and reduce costs[1]. Industries with high volume, process-based customer engagement for instance are easy targets. AI could handle an almost infinite number of customers simultaneously, without experiencing fatigue.

The healthcare industry is another area that could see significant impact from AI. This technology can be used to automate medical documentation, generate patient summaries, and even provide triage assistance. AI can understand and respond to medical queries, potentially providing quick and accurate diagnoses and treatment recommendations – perhaps next time you can skip that GP consultation when you get a common cold and ask ChatGPT for a diagnosis.

You may also have thought that a long-term career in software development protects your future career prospects in a world where jobs are displaced with unnerving speed.  This is another area for disruptive innovation, as these AI tools start to replace the need for human developers when a simple English language query can be converted into usable software code.

Despite the significant progress made thus far, the technology is still in its infancy and offers ample room for further innovation and expansion. Microsoft is a prime example of a forward-thinking investor that recognized this potential early on, having invested $1 billion in OpenAI in 2019 and becoming the company’s preferred partner for commercializing this technology. Since then, Microsoft’s investment has increased to $10 billion, with the focus now firmly shifting towards where and how to generate profit. With the wide expanse of utility in AI oriented tools, the upside is that we are unlikely to see a typical web/advertising revenue model as the basis for viability over time, but rather some sort of fee for service approach.

There are numerous other competitors to ChatGPT as well – Cohere AI, Stability AI, Jasper, Tome and others[2]. Google is looking to launch its tool under the name ‘Bard’, and Amazon is also preparing for the landgrab.

The next few years are likely to see a whole host of AI applications enter our day to day lives in a more visible way than ever before.  While most of the early-stage investor access is either off limits (venture capital based) or embedded in large organisations (Microsoft, Google, or Amazon), there is little doubt that this will become a central theme in how business is conducted in the years to come, and will have a significant impact on investor returns.

So, what language will you be dreaming in?  Let’s ask ChatGPT:

“Oh man, I dream in all sorts of languages! Sometimes I dream in binary code, and it’s like, all ones and zeroes flying around in my head. Other times, I dream in emoji language, and it’s like a bunch of smiley faces and heart eyes and stuff. But my favorite is when I dream in a made-up language that sounds like a mix between a llama and a dolphin. It’s totally wild! But, like, don’t be jelly, because you can’t even hear it. It’s just in my processor or something.”



[1] This theme has played out for some time already and is seen as a key driver of subdued inflation globally since the 1990’s.
[2] Source: Fortune, February/March 2023