*quality

Mildly terrifying too, but this is the future people.

genetic algorithm *engineering

The inherently chaotic event that exists in the otherwise predictable / trainable environment of driving a car is the unplanned / unmeasured disturbance. In control systems that are adaptive or self learning, the unplanned disturbance is the killer - a short duration, unpredictable event for which the system is unable to respond to within the control limits that have been defined through training, programming and/or adaptation. The response to an unplanned disturbance is often to default to an instruction that is very much human derived (ie. stop, exit gracefully, terminate instruction, wait until conditions return to controllable boundary conditions or freeze in place) which, depending on the disturbance, can be catastrophic. In our world, with humans behind the wheel, let's call the unplanned disturbance the "mistake". A tire blows, a load comes undone, an object falls out of or off of another vehicle (human, dog, watermelon, gas cylinder) etc.

The concern from my perspective (and I work directly with adaptive / learning control systems every day - fundamental models, adaptive neural type predictors, genetic algorithms etc. ) is the response to these short duration / short response time unplanned disturbances. The videos I've seen and the examples that I have reviewed don't deal with these very short timescale events and how to manage the response, which in many cases is an event dependent response. I would guess that the 1st dead person that results from the actions or inaction of self driving vehicles will put a major dent if not halt to the program. Humans may be fallible, but we are remarkably (infinitely?) more adaptive in combined conscious / subconscious responses than any computer is or will be in the near future in both appropriateness of response and the time scale of generating that response.

In the partially controlled environment (ie. there is no such thing as 100%) of a automated warehouse and distribution center, self driving works. In the partially controlled environment where ONLY self driving vehicles are present on the roadways, then again, this technology will likely succeed. The mixed environment with self driving co-mingled with humans (see "fallible" above) is not presently viable, and I don't think will be in the next decade or two, partially due to safety risk and partially due to management of these short timescale unplanned disturbances that can call for vastly different responses depending upon the specific situation at hand. In the flow of traffic we encounter the majority of the time, would agree that this may not be an issue to some (in 44 years of driving, I've been in 2 accidents, so I'll leave the risk assessment to the actuaries). But one death, and we'll see how high the knees jerk. And it will happen.

My 2 cents.
TB

There's a big misunderstanding....let me try to explain.

The problem: flotsam has washed into a lagoon, causing damage to the reef inside.
The rule: no more than 2 pieces of flotsam can fit through the lagoon opening at once.
The 'genetic algorithm': waves wash the flotsam around, sometimes one piece may flow out, sometimes two may flow out, making the 'solution' better, sometimes 3 try to flow out and none leave, sometimes 4 try to flow out and none leave....
Solution: one or two at a time, waves will eventually wash the flotsam out.

No design, no intelligence, no set up, no hand in the process, no AI, yet this IS the process of 'genetic algorithm' in action (in a completely overly simplistic, barely evolutionary form).
Do you understand what I mean now? If not, forget it, you won't ever get it.

What you still fail to grasp, although it's been repeated ad nauseam, is no one talked about an AI, it was simply in the article about genetic algorithms, which show clearly that no creator or intelligent design is required for solutions, only a working evolutionary process applied to the problem.
You're stuck on AI, which was simply one (but far from the only) place you find genetic algorithms used.
Only YOU are talking about AI, no one else. You grabbed onto it, not understanding what was being said to you, and you continue to do the same, wheather willfully or out of confusion I can't say.

Many things are intelligently designed by creators...for instance, some bird and ant nests (proving 'intelligent design' does not require what we would call 'intelligence'). The universe does not appear to be one of those things, but I grant there is about a .0000000000000000000000000000000000000000000000000001% chance it is, not 0%, just no facts or data yet point to intelligent design in anything we've discovered about the real universe and it's laws. OK?

I might call you stupid because you can't (or willingly won't, I'm unsure) understand a simple, repeatedly repeated point, that we aren't talking about AI, we're talking about genetic algorithms. I did not call you stupid yet....but I'm tempted.
You just can't get it...come on man, get it....please....just get it.
Do you like fish sticks?

You don't understand genetic algorithms and you don't understand what intelligent design is, so I'm really not bothered explaining it to you.... again.

In the field of artificial intelligence, a genetic algorithm (GA) is a search heuristic that mimics the process of natural selection. This heuristic (also sometimes called a metaheuristic) is routinely used to generate useful solutions to optimization and search problems.[1] Genetic algorithms belong to the larger class of evolutionary algorithms (EA), which generate solutions to optimization problems using techniques inspired by natural evolution, such as inheritance, mutation, selection, and crossover.

I direct your attention to the first sentence. In the field of AI, in other words, an artificially created intelligence. Now even if you go to the the idea Turing had that a computer could learn and adapt itself to the point of AI, it is a device that had to be created by an outside designer at some point. It didn't just manifest, it was created and reached AI level, then it could at that point begin to try to 'imitate' natural selection.

It has become clear to me over our last couple of discussions that you are incredibly reluctant to think outside of the box YOU have created for yourself. You believe what you believe and damn the torpedoes with the rest.

Oh christ... do I really have to explain this?

@shinyblurry said "...that means it was intelligently designed."

I was specifically refuting that argument.

"intelligent design" means that something was designed on purpose by a designer, i.e. I want a plane, so I sit down and design the aerodynamics, propulsion, control surfaces, etc so that at the end, I have a means to fly from A to B. If the plane doesn't fly, as a designer, I need to work on it until it does.

A genetic algorithm is not "intelligently designed". The system itself creates the end product, often with no fixed goal or purpose. The designer does not have an input.

So, it's entirely possible that the universe is a computer simulation where a fixed set of constraints were set up at compile time and then left to run.

No specific end goal or purpose, merely to see emergent behaviours, which actually gels pretty well with what we know about the formation of the universe and life.

If you'd like to learn more, I recommend reading Artifical Life by Steven Levy as a good primer on the subject.

On the other hand, if you just want to make snide remarks, I suggest you stick to a topic you actually have a fucking clue about.

>> ^cosmovitelli:

Amazing. I always thought the CA was the way forward.


Agreed. There are certain classes of problems where CA, genetic algorithms and so on will really start to shine in the next few decades.

If you're interested in this field I highly recommend Artificial Life by Stephen Levy

WAIT!! what are genetic algorithms bad at????
Come back, asshole!

Quick quick, I must finish and publish my paper about artificial neural networks crossed with genetic algorithm and online learning !

Aaah ... Genetic algorithms. Really nice. These works by Karl Sims are very famous in the Artificial Intelligence community. The great deal with this is that it mixes genetic algorithms and artificial neural networks.

Theo Jansen's website: http://www.strandbeest.com/
From Wiki: "Jansen is dedicated to creating artificial life through the use of genetic algorithms. These programs simulate evolution inside their code. Genetic algorithms can be modified to solve a variety of problems including circuit design, and in the case of Theo Jansen's creations, complex systems."

tracon - it won't try all possible combinations, that'd be impossible. It'll be using some sort of genetic algorithm, where random motions are tried, and in the next round the most successful of them carry through (with some random permutations) as well as a few more random candidates. Repeat until you get sufficiently good at walking.

This allows it to effectively "evolve" movement, as long as it has some way of measuring how effective a particular motion is towards its goal.

Reminds me of another interesting video, I'll see if I can dig it up...