top of page
Search

AI’s Humble Beginnings

As AI has become widely accessible in the last two years, we often describe it as the most significant technological transformation of this decade, implying it's new. And yet its origins are surprisingly much older.

The timeline goes back to the 1940s when electronic computers were first invented. The first shown AI system is ‘Theseus’, a robotic mouse from 1950 invented by Claude Shannon. It became a field in its own right at a 1956 Dartmouth Conference, which brought together some of the brightest minds of the time to explore the possibility of machines simulating human intelligence, which lay the groundwork for one of the most transformative fields in modern technology. And yet, only as recent as 10 years ago, no machine could reliably provide language or image recognition at a human level.


AI systems that rely on machine learning require training; algorithms and the input data for the training. As the training computation has increased, AI systems have increased in power and their capacity to do more complex tasks, which is why it appears AI is a relatively new concept in the last few years.


Sixty-eight (68) years is a long time in the making for AI to become mainstream and yet now we are seeing the expectation to adopt, integrate and leverage effectively, Generative AI, move at lightening speed.


For marketing and sales functions, research by McKinsey demonstrates the expectation of leadership is that the most significant gains from deployment of Generative AI technology will be in their marketing and sales functions, even higher in high tech industries.


So if the last two years are any indication, what does the future of AI look like for businesses?


Not too dissimilar to the last technological innovation of cloud, we will see:


  1. Improved accessibility and wider adoption - continued advancement in AI will make the use of it more pervasive in everyday life. Think personalised learning experiences or enhanced assisted technologies for the vision or hearing impaired.

  2. Specialisation and efficiency - we are already seeing the development and use of SLM's that are tailored for specific tasks or industries; improving accuracy and relevance as they are trained on specific data sources and requiring less computational power so they produce faster results.

  3. Commoditisation - Just like cloud, LLM's enable faster innovation and development and as more players enter the market, costs decrease, contributing to the commodisation of AI. Building AI technology requires a nuanced understanding of how to craft prompts (prompt engineering) that produce the desired outputs. As more people use LLM's and their collective knowledge around prompt engineering improves, we will see commodisation in the tools themselves.

  4. Ethical and Responsible AI - at the time of writing, this is a highly unregulated space. With more users, there will be even greater scrutiny and discussion around the ethical implications of AI. Governments worldwide are actively working to regulate AI technologies, including LLMs, with varying degrees of focus across ethical use, governance for high-risk AI applications, social order and moral standards and transparency.


AI, though often perceived as a recent innovation, has roots that stretch back to the 1950s. Today, AI has become integral to many industries, with Generative AI leading a significant transformation in marketing and sales functions. The rapid advancements in AI, particularly in the last two years, mirror past technological shifts like the adoption of cloud computing, paving the way for broader accessibility, specialisation and ultimately commoditisation. For businesses, the future promises a landscape where AI not only enhances efficiency and effectiveness but also demands careful consideration of its broader societal impacts.


Sources:



 
 
 

Opmerkingen

bottom of page