One of the most useful developments in AI is in oncology. AI tech is learning fast how to identify cancer from imaging which has perhaps been missed by human eyes.
There is also some development in AI personalising/optimising chemotherapy.
Absolutely, any kind of vision task and vision classification AIs should be better at than humans (and are already for many applications). Medicines, chemicals, materials - all will benefit from AI.
The big unlock will come when AI is better/faster at developing AI than humans.
It’s also a question for me whether AI is currently learning slowly because we are teaching only via text and images and whether if we embody AI into a robot with sensors it can learn as fast (if not faster) than a human. After all, humans by age 5 can already do a lot and understand a lot.
I would expect that an embodied AI should be able to learn much faster than a human as it wouldn’t need to sleep and could also have a shortcut by having all human knowledge already downloaded so it can cross-reference all its experiences with this knowledge. e.g. when it learns to walk, it can do so with the full framework of mathematics and newtonian mechanics in the background.
The philosophical question is are we ready as a species for AI? The more our machines become like us the more we will merge and could this be the next step in our evolution?
I listen to a hacking/cyber-security podcast and the technology exists now to clone a person’s voice and mimic it so well, it’s almost impossible to tell the difference between the AI voice and the real person talking - pauses and inflections are all there to make it more realistic.
It’s used for phone call Penetration testing but also by criminals wanting information.
Such technology was already used for crime. One criminal created a video call where all people on the call except the target were fake. The fake meeting had some discussions ending up with the boss authorising payment to the criminal which the target duly processed.
Who knows, maybe that $90k represents 4 jobs that were lost over the span of 2 years. Now replaced with 1 guy that is 250 times more time-efficient with AI.
I’ sure there is some skill involved too and it’s not that anyone could do it.
We don’t quibble when someone comes to repair something, plugs in their laptop, runs a diagnostics program which takes seconds and points out the problem part which then can quickly be changed.
The repair person doesn’t charge less because his computer solved the problem.
Ultimately we’re happy that the problem was solved efficiently and we got the result we wanted.
That’s the same with the Playing card artwork. The result is what is important - not the method in getting there - unless the hand-made aspect is a cherished part of the product and is a selling point.
I’m not arguing against what he is doing, only that if this increases, then we will need to find a lot more jobs as one person is able to replace several workers which will increase unemployment.
e.g. if he replaced a team of 3 designers, maybe then that is an office which has a receptionist, payroll, cleaner, accountant etc.
Or moving people to more right-brain type jobs. Here’s a pretty good article in Forbes which describes how we may still have one up on AI for the foreseeable future, or where humans and AI can blend more seamlessly.
And what happens to the “computer illiterate?” A phrase I heard in the eighties by the way. Does society split into two separate tribes? I see them almost every day, as long as they can find the correct buttons to push all is well, but they go to pieces the moment you ask them to do something other than press button “one”
They learn. If my 9 year old can learn to use Copilot on my laptop within 5 min, it shouldn’t be too hard for anyone. Its such a powerful tool already that there’s no excuse for not using, experimenting and playing with it already.
Having said that, i somehow prefer chatgpt vs copilot
The right side of the brain is responsible for controlling emotions, motivation and creativity while the left side controlled functions include language, reasoning and logic. So I think right brain type professions are probably more amenable to computer illiterates.
You’ve reached the current usage cap for GPT-4. You can continue with the default model now, or try again later. Learn more
which is strange, because I use free version 3.5.
did someone have similar situations already?
I really don’t want to write alone email to the insurance company on german