Sam Altman, CEO of Open AI and father of ChatGPT, recently discussed key details about GPT-5 and the future of AI, including the importance of multimodality and the potential for video representation in AI models.

Incorporating images and video in the interactive process of providing input and responding to questions will be an important milestone in future AI applications. This is already available in ChatGPT-4, and it will be further improved.

Sam Altman also explores the shift towards personalized AI agents, the increase in data elaboration in terms of reasoning ability, and the 'increased intelligence' of the model.

Future models will be able to increase their knowledge in real-time from daily updates coming from the internet. Current models like GPT-4, instead, need pre-training.

How AI could impact future jobs

The ongoing evolution of AI will strongly impact human jobs that could be replaced by that easily accessible ‘intelligence’ - fast, reliable and inexpensive.

This covers several white-collar jobs, from accountants and legal professions to various engineering jobs, including software development and basic coding.

The last example is significant: if AI is ready to replace junior programmers now, this could create a generational gap in the human workforce among developers. What will happen to the next generations of human programmers?
Is it possible that in 10-20 years, we could have very few human programmers at all?
What will this entail?

This AI could obviously be applied to smarter robots as well. These smarter robots could impact many blue-collar jobs as well, and we are already seeing some of this with the first self-driving cars, smart warehouse logistics, and automation in construction.

This AI is able to give better brains to robots and, most importantly, a smarter interface to communicate with humans. In other terms, robots that not only react by responding to buttons, graphical interfaces, or maybe joysticks but are also able to communicate in both input and output, talking in the same way as a human would.

Do We Need an Increase in Productivity?

We are looking for a significant improvement in productivity from AI, and that is right; however, giving people more powerful tools not only allows them to work faster but, more importantly, it will enable them to do qualitatively different things.

It is not just about 'doing more stuff' at the same time; it is also about 'doing stuff at another level of complexity'.

To use an example: humans can 'run,' but thanks to a tool like a bicycle, they can learn to 'run' at another level just by pedaling. Some people will say: 'running' is human - 'cycling' is … artificial? Or is it an augmentation of 'running'?
How do you comment on that?

The existence of that AI will force us to adapt.
This is not something new in the history of humanity.

Several technological revolutions have historically been absorbed well by society. Many jobs that people have done have massively changed over a couple of generations.

The fact is that each technological revolution has gotten faster, and this will be the fastest by far.

This is the difficult part: the speed of these new changes has reached a level at which society and the labor market will have to change so fast that this is unprecedented and could result in a more challenging situation to absorb for human society.

The same experience was decades ago for blue-collar jobs and manual labor taken by machines and industrial robots. Now, this is happening to 'white-collar jobs' and several medium-level skill jobs. However, current AI enhancements will impact the new generations of physical jobs too because a new level of robotics is going to be available. Self-driving cars, robots managing logistics within a warehouse, or many other examples of smarter and smarter robots can perform not just more tasks but more complex jobs entirely.

What Could Be the Most Revolutionary Difference of Future Tools and Machines?

The way in which humans communicate with these machines or devices could be the most revolutionary aspect. The concept of an advanced interface, where the interaction is conducted by 'talking' in the human way instead of clicking buttons, using monitors, and tapping into graphical interfaces, could take everything to another level.

Indeed, we could change the paradigm of interfacing with our devices and computers. No more opening a browser and clicking around menus and buttons with mice and keyboards, but simply using spoken interaction in the same way it could be done with another human.

So, we could ask our computer: 'What are my most important emails today?', 'Can you reply to the message from John?', 'Can you book an Uber for 10:00 AM tomorrow at my home address?'

The list of these examples could go on, covering any sort of business tasks as well.

What Happens to Human Jobs When AI Surpasses a Human's Abilities?

A question immediately arises: What about the role of humans when AI surpasses a human's abilities? The answer to this probably lies in another mindset shift.

If we continue to stick to the concept of 'old' jobs, these will be replaced. However, considering human jobs where AI is present as a smart collaborator means the role of humans will shift to higher levels of management, coordination, and creativity.

Humanity will never run out of problems to solve, and the difficulty of these problems is virtually infinite.

There are no limitations about that. When some jobs are taken over by AI, humans must level up to another level. And that doesn’t need to be exclusively a 'job'. It could be creativity and entertainment too.

That will definitely increase, and indeed, it is exactly what has happened in the last decades.
The recreational and entertaining industry are sectors that have registered a great increase in the last decades.

Life in 1920 was much more 'sweat' and less 'Netflix and video games,' just to put it simply. How many new jobs were created by that shift? Probably no boy born in 1920 would have ever imagined making his living as a film actor in 1920, even though that technology had appeared several decades earlier.

Finally, advanced AI could provide substantial support to research in science, physics, and medicine in the long run.

It is already a fact, for example, that protein engineering and advanced biology are more often done by computers than humans nowadays.

So What?

Humans will be challenged to solve higher and higher problems, and we will probably make reality some fantasies pictured by visionaries like Gene Roddenberry (the father of Star Trek).

Taking inspiration from science fiction, we have seen all kinds of scenarios, starting from the ones from authors like Isaac Asimov or Arthur C. Clarke and so on. Then, we have also seen bad scenarios.

The game is set up, and it is up to us humans to play it well, using all the resources in our hands to build our future society. That definitely will include the smart use of Artificial Intelligence!

Future generations will live in a society where artificial and biological intelligence will collaborate for evolution.