What if the brain can interface with the device on a hierarchical chain instead of equally? It may be able to delegate computing tasks to the interface while retaining everything that is related with consciousness within the biological part.

You would have to see what is the minimum biological brain size where a person is alive and functioning well, then connect the interface and show that - with the device attached - you can reduce the biological brain to a size that would otherwise be impossible and still have them retain consciousness. Then it must be shared with the machine.

if the mods think its a problem, the mods will take care of it. just tell them.

its their sub, not yours, not mine.

We’re not as smart as we think. We can hardly increase our intelligence. We aren’t all that general either.

That's not a 1:1 copy of the human hand. Real human hands are covered with skin and have lots of touch sensors to support manipulation. This video is about an animatronics hand for setting up a show.

Id love to see more about this experiment and more like it. It’s fascinating. In my (limited) opinion I think we will find out that consciousness itself is an abstraction. Like how a chair is a chair because we say it is.

If computing cost for AI is really high, it might be more effcient for humans to do some jobs still

I feel like half the blame is on the survey itself, which apparently had all kinds of weird/arbitrary questions and asked for probabilities framed in 3 sets: 10 years, 20 years, and 50 years.

When you ask someone to put different probabilities into 3 timeframes like this, they're going to be biased to lowering at least the first one just to show an increasing probability over time, with the first being 10 years away and the last being 50 it makes sense that every time they do the survey their result is going to make it seem like everything beyond what is already public and well known is going to take forever to happen.

For the second part of the blame, I'll cite this example:

"AI Writes a novel or short story good enough to make it to the New York Times best-seller list."
"6% probability within - 50 years"

not sure who answered that, but they're probably an "expert"
just sayin'

Breaking news: Bernie says something idiotic again

What happens when bots are good sex mates? Give you a good blow so you don't have the urge to chase.

Machine learning

These things always take today’s tech and put it on a linear path. The key inflection is AI that writes code for AI and AI that designs chips for AI. We passed that mark decades earlier than this prediction expected. We are now past the event horizon and in the gravity well.

"I feel like if people could be young and beautiful it would solve a lot of problems in this world"

This is the weirdest take. How does people being beautiful improve anything?

Imagine asking a pre-Bronze Age hunter-gatherer what to expect from the year 2023, and comparing it with what really happened. Then apply the analogy to today, only moreso.

I wasn’t challenging your point. More of a reply out of curiosity.

What’s ML

ChatGPT is a turning point, I think, because the average person can see firsthand it's effects. people pulled the fire alarm back in the GPT-3 days (only 3 years ago, if you believe that) but now even Vox is writing about AI alignment issues

The thing they're predicting has nothing to do with anything related to GPT.

Was gonna say, I followed the sub because of stuff like that. If you don't like it, scroll past it. Or contribute your own material on a different subject or topic.

Entertainment

One transistor is loosely analogous to one neuron but definitely not the same.

But then to have enough transistors to simulate all the neurons needed you will need to consider quantum effects to design them, nvm.

diamonds man.. fucking stone.

human enterprise is built on lies.

I think you would like this one: https://open.substack.com/pub/doxa/p/machine-learning-and-deflationary

Basically observes that all prices are headed towards zero.

No more overpopulation because AGI will supply low-cost mates.

What do you mean by that

Surveys don't predict technology. And who knows if any of these people are working towards AGI.

If you want technological predictions, you need to look at the information put out by people trying to make those predictions, which involve tracking trends in requirements such as cost of computation, cost of energy, funding rates, scaling efficacy, etc.

The original post talks about AGI, not ASI or technological singularity.