A featured contribution from Leadership Perspectives: a curated forum reserved for leaders nominated by our subscribers and vetted by the Construction Tech Review Advisory Board.
Head of Artificial Intelligence and Machine Learning Dept. at Gruppo MOL
The Evolution of AI in the Corporate World
Everyone always says we need to look into the past to plan for a good future.
I think this is not always true, simply because the issues we face today were not there in the past. But one thing is true: looking into the past will at least give you an understanding of the velocity at which events evolve. This concept is known to people working in agile contexts.
When I started working in the artificial intelligence (AI) and machine learning (ML) ecosystem in 2017, things were different in terms of both technology and the readiness of the business stakeholders. My team not only had to explain every single project rationale but also had to dive deep into the details of every feature used by estimating the impact it would have on production systems. Nowadays, things are more straightforward due to the significant impact of LLMs like GPT and Bard. On the one hand, corporate employees are no longer scared of AI and are generally very open to adopting those solutions immediately after viewing a simple MVP of their use case. On the other hand, their expectation level was raised to the stars: "Can ChatGPT do all the work?"
This introduction is probably oversimplified, but it should give a fair image of the drastic change in people's mentality and understanding of AI-related technology. In a previous article, I tried elaborating on some concepts related to AI replacing humans, and, at that time, I underestimated the velocity of technology progress; I couldn't even imagine that in just a few years, NLP and NLU's maturity level could generate something like GPT-ish models.
To bring some reality to the table, I can elaborate on some tasks that I've seen (in a business context and not just on the research desk) evolve from pure manual to almost completely automate in a time shorter than I could forecast:
● Data entry: Many back-office tasks were performed by humans who copy text from website to website, digital documents to applications, or paper documents to applications.
The first two examples were automated simply using RPA; the latter, which is more complex, now depends on OCR's quality more than the underneath computer vision model (or NLU or other different ML technology). Document classification is a similar (maybe easier) problem already solved by such algorithms. This is the area covered by Back Office teams where machines reach human-level performances.
● Summarising key concepts in most spoken languages after understanding and extracting key concepts from human discussions (even code): This area is covered by executive assistants, project managers and technical and copywriters. We could extend to brokers and client-facing people who explain contracts and product features. Any co-pilot technology has reached a very good level of maturity.
● Dealing with end-users with tasks like information collection, subscription process, signature and document verification: This is an area covered by sales assistant, hotline support and post-sales customer support. Most of the above tasks can be performed by LLMs at the cost of providing a correct and detailed prompt.
● Writing code in a widely used programming language provides a detailed description prompt. Again, this technology is available and integrated into most coding UI.
Are the results good enough to be used for production?
It depends on how critical the task is and how much error someone can sustain before impacting revenue. As a matter of fact, all AIs make mistakes: if your process is very sensible to errors, then you still need a human in the loop that can correct the machine (and allow it to learn from mistakes).
On the one hand, corporate employees are no longer scared of AI and are generally very open to adopting those solutions immediately after viewing a simple MVP of their use case. On the other hand, their expectation level was raised to the stars: "Can ChatGPT do all the work?
Machines need humans to learn, and business needs humans to ensure everything runs smoothly, but what will happen in the future?
One of LLM stated that: "In summary, the future impact of AI on employment will be a complex interplay of job displacement, job transformation, skill evolution and the emergence of new opportunities. How society manages this transition, through education, policy and innovation, will play a crucial role in determining whether AI leads to a more prosperous and equitable job market."
On the same side, a human forecast states that we'll have a modest socioeconomic impact by 2030. By 2035, this impact will be significant and will not stop. The evolution will not be limited to the Back Office tasks but will extend to more physical jobs that involve movement, maintenance, 3D vision, medical and diagnostic. At a later stage, machines will be able to fully imitate humans in physical movement and creativity, opening issues like intellectual property and legislation.
We're getting to the point - humans must be prepared to evolve their roles by leveraging machine capabilities instead of fearing and rejecting them. Progress (and business) cannot be stopped.
