.png)
The future of AI: Preparing for what’s next
- lorenaflorian0
- 4 days ago
- 6 min read
Updated: 3 days ago
In recent months, AI has been an increasing as a buzz word that everyone seems to be talking about. You can sit in a cafeteria or office space and catch two conversations on the topic of AI. And with several professionals offering AI literacy course, it seems that everyone has the assumption of AI to be the next big thing that will disrupt business processes.
With the 2025 AIPM conference featuring predominantly the topic of artificial intelligence, it seems that the hype around the technology will continue to grow as the possibilities of AI continue to explode.
Because the capacity of technological innovation is highly uncertain and volatile, it is also important to recognise the uncertainty that technology poses. Back in late 1900’s, popular fictions have made guesses on what the future would look like, we made guesses of flying cars (Back to the Future II) and personal robot butlers (The Jetsons) that as of today still seem so far away. Yet some of these guesses are correct and sometimes surpassed such as the emergence of artificial intelligence and video calls in 2001’s Space Odyssey!
Although it has become a fun thought experiment on what AI would look like from today, it’s also important to respect the explosive uncertainty these technologies are capable of and ground ourselves in broader philosophies and perspectives that are made to last. With insights drawn from conferences and industry experts, here are useful perspectives you may adopt to respond to these emerging trends.
(in 1989, Back to the Future predicted flying cars would be the norm in 2010. We have yet to reach that point but we have experienced massive shifts in technology, namely Artificial Intelligence!)
AI’s unpredictable and explosive innovations
In early 2000’s the concept of artificial intelligence was mostly reserved for science fictions and wishful thinking.
Yet in 2018, OpenAI released the first GPT 1 model, an immensely unimpressive model which often results in confusing and outright wrong outputs. Certainly, unviable for commercial use.
2 years later, ChatGPT-3 was released with strong performance in coding, writing and conversations though it still struggled in accuracy and massive hallucinations.
Three years later, the GPT-5 was released featuring enhanced memory, personalisation, safeguards and tools utility with higher market viability.
In recent years, businesses are rapidly adopting Co-Pilot to automate and augment tasks such as idea generation, filling out paperwork and even creating an entire PowerPoint presentation. In some advanced AI models such as ChatGPT-5 Pro, they can even run their own tests and verify the papers for academics.
In under 10 years, AI progression exploded from a clumsy pattern-recognition algorithm to a commercially viable product that can replicate human intelligence convincingly. Whether AI progression will slow down or speed up, it is undeniable that it is making a steady pace into increasing capabilities and complexity.
Prompt: | Python code to find the smallest factor of a number: |
Gpt-1 | the lack of a body in the room before me. after several moments of silence, he spoke again. " you are my daughter. the two of us are one. and in time you will |
Gpt-2 | let p = &[ 5 - 3 ] => ( 1, 2, 3 ) The example above uses the "P" module to print the smallest factor of the number: |
Gpt-3 davinci | import math def lfact ( n ): factors = [ 1 ] factors . append ( n ) while n % factor <> 0: factor = factor * 2 - 1 while int ( factor ) > n % factor : factors . append ( factor ) return factors What the code does: Imports math module for math functions. [..] |
Gpt-4 | Here is a simple python code to find the smallest factor of a number. ```python def find_smallest_factor(n): if n <= 1: return None for i in range(2, n): if n % i == 0: return i return n [..] |
Gpt-5 | import math def smallest_factor(n: int) -> int | None: """ Return the smallest integer factor of n that is >= 2. - If n <= 1, return None (undefined / no factor >= 2). - If n is prime, return n. - Works for negative n by using its absolute value for factorization. """ if not isinstance(n, int): raise TypeError("n must be an integer")
if n <= 1: return None n = abs(n) # Check divisibility by 2 if n % 2 == 0: return 2 # Check odd divisors up to sqrt(n) limit = math.isqrt(n) for i in range(3, limit + 1, 2): if n % i == 0: return i # If no divisor found, n is prime return n |
As seen in the table, the first output was completely irrelevant, the second was not coded in python, third was invalid, fourth is correct but not optimal but the fifth not only correct but breaks it down into something more coherent and readable. All advancements were made under ten years from incoherent jumble to something equal to and even better than human output. It is possible that the role of curation will still be relevant to the coming years and those with specialised knowledge are required to verify the plausibility of each output, yet as the models grow in increasing sophistication, the future is still uncertain.
Another example of rapid AI development. The 2023 stable diffusion model was although coherent, had many mistakes and distortions when generating videos. Fast forward two years, Google’s release of Google Veo 3 marks a benchmark in AI generation with accurate video generation nearly indistinguishable from reality. In under five years, advertising and other services that require video recording can be cost-efficient and fast with the uprising of AI technology.
Not everyone will be on board with change
During 1811, there was a protest movement held by “luddites” consisting of skilled craftsmen that sabotaged technologies such as knitting frames and looms that were seen as a threat to undermine wages, replacing skilled workers with children and degrading work conditions.
Although we won’t have worker marching to data centres with hammer and pickaxes anytime soon, the spirit of resistance to technologies still ring true.
People will be resistant to changes in well-established workflow, especially when there is not a good enough reason to make these changes. This is why managers especially project managers that aim to increase the efficiency on the output of the project, whether it be improving cost, time or quality of the project, an appropriate stakeholder consultation and change management strategy is required before any project is executed.
Or else, technology innovations that threaten the competitive positioning of an organisation with inefficient change management will lead to businesses rapidly falling behind from their competitors.
(The introduction to filmless digital cameras were met with resistance to the senior directors at Kodak. Their inability to be on board with innovation and lack of process to leverage emerging technologies has led to the company’s bankruptcy as competitors rushed to capture the value of the technology.)
(People vested their confidence in soft-skill positions to be AI-proof, yet even PR positions and trust-building roles are at threat to the emerging trend of AI spokesperson)
Need for lifelong learning
After this, the crucial question arises that being, “Then what is AI-proof? How can me and my company survive the impact of emerging AI trends?”
By the time this article releases, and in a couple of weeks, it is highly likely that the technologies mentioned have already been surpassed with newer, accurate and sophisticated AI models that would be far more indistinguishable from those of humans and reality itself. And the simple answer is assuming that nothing is AI-proof and being vigilant, open and adaptable to emerging technology.
The Japanese philosophy “Kaizen” quite literally means “change” and “good” which sets out the core philosophy of continuous improvement. Although the AI will continue to explode in its capabilities, our ability to analyse them and adopting them to our workflow to increase productivity will continue to improve once we identify the need for it.
Conclusion
AI will continue to progress at a rapid pace in these coming years and adaptation for certain industry will soon become the choice between distinguishing yourself from peers or fall behind. It opens the possibility of the wonderful possibilities of AI, and because we can never know for certain what the future will bring, adopting vigilance for emerging trends and having the skill to adopt it becomes valuable and perhaps the norm in modern workforce.
Comments