Demystifying Neural Networks: Teslas are (probably) not alive, but that’s OK! (Part 6)

0

We have many guesses about consciousness

This secret of consciousness leaves a lot of room for guesswork and sometimes even educated guesswork.

Materialists think that we just have to study the brain harder and that we can find out what consciousness is by reducing it to matter. Unlike Galileo, the materialists think that the subjects that Galileo limited science to (size, shape, location, and movement) are basically all that the universe has to offer, so we just have to keep trying, the material To find thing that creates consciousness in the brain. Weak reductionists are materialists who believe that consciousness arises in the brain, but that due to differences in perception of consciousness and neuroscience, we probably cannot use science to find the connection.

Dualists believe that the mind exists outside the brain, but there is great disagreement among dualists as to how this works. This can border on spirituality and religion as this reflects the beliefs of people of many religions. The idea that a person’s “mind” exists outside of the body goes well with this.

Panpsychists believe that there is a bit of mind in all matter, and that if you bring together enough complex matter it will lead to awareness at the level that people experience it. Goff (the author I cited above) tends to fall into this camp, and explains at length why, in his opinion, it is currently the most logical and therefore worth further interdisciplinary investigation.

Others think that the universe is conscious or that the brain is somehow connected to the “fabric of the universe”. This area of ​​thought tends to attract many New Age spiritualists, mystics, and others, but there are scholars who wish to separate their study of this subject from the belief systems of others.

I have my own opinion as to which of these is right, and I will not lie and say that my spiritual views do not affect my position. But I won’t go into that here. We all have our beliefs and opinions on this unresolved matter, and mine is probably no better than yours. However, this could be a fun topic to explore further in the comments!

Why discuss all this spiritual mumbo jumbo at all?

The only reason I mention all of this is to solidly separate and separate from the harder facts the mystique and spirituality that invades discussions about consciousness and AI. When it comes to the latest in consciousness research, the field is currently very open and unknown. There is a lot of information out there and a lot of people making educated guesswork and it’s an exciting time.

There are also many people who claim to have all of these answers already. Religious leaders, quantum mystics, and even people who basically worship Elon Musk all have very solid ideas about how this works (or not). I’ve seen people absolutely sure their Tesla is alive, saved their family from a situation it wasn’t created out of, and more.

Everyone is entitled to their beliefs, of course, but we have seen no evidence that Tesla’s vehicles are imbued with awareness, and we would be foolish to count on this hypothetical awareness of safety. We need to make sure that there are other plans in place in case they’re not alive.

A lack of awareness does not mean that self-driving cars are impossible

Without a breakthrough discovery or the accidental creation of consciousness in a computer (given the unknowns, we could actually go wrong), we shouldn’t rely on Tesla to develop a living computer for its cars. That would be amazing because the car’s computer could rely on its training when it was good enough and improvise when it encounters gaps in its training just like we do, but that is, as far as we can tell, unreachable now.

Without the ability to improvise, adapt, and synthesize new ideas, Tesla’s computers must instead run programs like any other computer ever built. She can fully rely on how applicable her training data is to the real world. The more accurate the training data and the better the training, the fewer edge cases there are.

It is basically impossible to ever achieve 100 percent reliability because the world is simply too complex to ever collect enough training data to cover it all. But Tesla can come up with a system that is almost 100%, and that’s good enough.

If Tesla’s FSD Neural Network is applicable to 99.9% of trips, it means the computer will make about four hours of errors for a year of driving this year, or about a minute for every 24 hours of driving. Say you drive 2 hours a day, that’s somewhere on the order of 5 seconds of erratic driving every day, and that’s going to happen by chance.

We all know that a lot can happen on the road in 5 seconds and that could easily kill you, so people wouldn’t be happy at all with a system that is only 99.9% good.

Fortunately, Elon Musk has said that they are aiming for “eight nines” in reliability. That’s 99.999999% reliable, or six more nines after the 99%. Fantastic!

But what does that mean? It is about 1/3 of 1 second per driving year. Remember, we’re talking about 365 days of full driving, 24 hours a day (which no one really does). If you drove two hours a day, that 1/3 of a second would happen every 12 years of operation.

Now we’re talking about a system that human passengers can seriously trust.

Achieving this, however, is not going to be that easy. When it comes to computer services, including those for hospitals where lives could sometimes be at stake, the question still arises as to whether it is worth getting even “five nines” (99.999%). Many non-essential services promise an availability of around 99.0%. Sure, we would be happy if all of the things we use were “eight nine” reliable, but we are usually not willing to bear the costs involved.

I do not pretend to know what the cost of reaching Eight Nines will be, but it will clearly be a huge cost that will take a lot of work to achieve and a lot of work to maintain that level of confidence. I see Tesla working seriously on this front, so I’m cautiously optimistic about their chances of success.

To make it easier to navigate this long series of articles, here are links to the others:

Part 1: Why Computers Only Process Numbers
Part 2: False Confidence in Math
Part 3: computers only run programs
Part 4: How Neural Networks Really Work
Part 5: What Artificial Neural Networks Can’t Do

Featured Image: Screenshot of Tesla’s AI Day.

Do you value the originality of CleanTechnica? Consider becoming a CleanTechnica member, supporter, technician, or ambassador – or a sponsor of Patreon.


advertising



Do you have a tip for CleanTechnica, would you like to advertise or would you like to suggest a guest for our CleanTech Talk Podcast? Contact us here.


Source link

Share.

Leave A Reply