So where are we with AI?
“ChatGPT has brought AI into the collective consciousness of people”
Everyone is trying to get their head around artificial intelligence (AI).
What exactly is it?
Do I need to use it? (Yes!! :)
How can I use it?
What does it mean for me?
What does it mean for my kids?
What does it mean for our future?
What does it mean for humanity?
Trying to answer these questions for yourself can send you into a deep and dark rabbit hole.
Ask me, I have been there many times.
Type in Google “What is the future of AI” and you will get over 2 million results.
Where do you even begin!?
To form my own opinions, I seem to always end up choosing to listen to Sam Altman, the CEO of OpenAI, the company that built ChatGPT.
I make it a point to listen to his talks, rather than read or listen to someone else who is analysing his talks.
I like how he speaks. I like that he is optimistic, but also realistic.
I like that he technically understands how the large language models that are building AI systems, work.
And whenever I listen to him, I feel his heart is in the right place. Super important when you are a key player responsible for building frameworks of AI that will eventually be super intelligence.
I also find it reassuring that although he understands the potential doom ‘super intelligence’ can cause, he is not afraid of it.
Of course, no-one yet knows what Artificial General Intelligence (AGI) will bring, and what ‘super intelligence’ means, but if anyone can fathom its potential today with the technical knowledge to back it, it’s Sam Altman.
There are many people who are positioned as ‘experts’, but I don’t know enough about them to want to listen to them.
Also, I’m not sure if they know what’s going on behind the scenes, purely because unlike Altman, they are not building the systems themselves.
I recently listened to a Wall Street Journal interview with Sam Altman and his CTO Mira Murati. I highly recommend it if you are looking to understand where we are and where we are heading.
What stood out for me from the WSJ interview on AI and its future were these points:
The development of AI is fast and new baselines are being established quickly, so the goal posts of AI development keep moving.
However, we are still quite far away from ‘artificial general intelligence’ (AGI).
Building these systems are extremely expensive, and a lot more money (think billions of $$$$$$$) will be required to get them in place.
Altman doesn’t worry about things like disinformation and deep fakes, because we are beginning to recognise them; a bit like when photoshop came out - it’s not hard to tell when a photo is airbrushed or photoshopped.
The biggest limitations AI has at the moment (think GPT) is 1.) its limited ability to reason 2.) its lack of reliability and consistency of information output (for example: ask ChatGPT something 10,000 times and you will get different answers with different information.)
Also, at the moment AI has no judgement, empathy or creativity and he believes that in that realm, it will never likely supersede humans.
He doesn’t fear the job market being disrupted because it’s a sign of progress, just like any revolution.
Putting a tool like ChatGPT into the hands of the public is a step towards equalising access to information and giving everyone a tool to work better and faster.
The most important point that both Sam and Mira make in this interview is:
“ChatGPT has brought AI into the collective consciousness of people.”
The only way we as a society can start learning the capacity of AI is by actually using it.
By using it, we will get to know it better and adapt to it, and similarly the tools will get to know us better and adapt.
There is debate over whether tools like ChatGPT should have been shared with the public whilst not quite ready.
Altman believes that there is no other or better way to introduce AI into the world. It’s coming into the world at an unprecedented rate, and will soon be part of our everyday lives. The earlier we know about it and start using it, the better.
Working on building this kind of intelligence in a lab would be like grooming a ‘perfect’ child behind closed doors in isolation, and then releasing him into the world when he is 18.
Can you imagine how messed up that would be both for the child and for the environment he is released into?
Releasing ChatGPT into the world for free, for everyone to access is the best thing Open AI did. Not only for the tool, but also for us as a society.
It is enabling us to catch pace with this new technology that is going to be part of our every day lives sooner than we know it.
Only if everyone has access to AI, and is using it, can we collectively begin to understand it.
Only if we understand what we are dealing with, can we start thinking about the ethics of using it, and the challenges we might face as a society.
Only then, we can tackle them collectively.
What then do we need to worry about?
The biggest fear that Sam Altman has is when AI will be able to influence or persuade individuals.
He feels that we will know when a robot or a machine is talking to us; we already do so there is no danger in that. But when that machine has the skill to reason and persuade us, is what we should be fearing. He refers to it as the fear of “individualised persuasion”.
Although I do not entirely understand what that means, I’m guessing it means a kind of manipulation by machines is what he is fearing.
However, he also hopes that by working as a collective society on the challenges ahead of us, we will be able to put the ethical regulations in place to avoid the potential destruction that AI can cause.
Ultimately, AI is a tool. What it will be is what we will make of it.
As Altman puts it, “people remain the architects of the future, not one AGI in the sky.”
Right now, what we have is an algorithm that can truly learn and it gets predictably better with scale.
Tools like ChatGPT can free us from the drudgery of work, giving us back hours of time in the process.
AI is going to become a part of society, whether we like it or not.
Let’s take the effort to understand it, use it properly, and embrace its potential for good so we can make the most of it for both us and future generations.
PS: I’m hosting a live workshop on March 7 on “How to train ChatGPT to be your personal assistant”. Since you are readers of my newsletter, use code: GPTWMC at check out to get 20% off till February 27 :)