| Not researched |
Factual |
15 Jul 2018 |
Brooks traces traditional robotics, as an approach to AI, to Donald Pieper's 1968 work 'The Kinematics of Manipulators Under Computer Control' at the Stanford Artificial Intelligence Laboratory (SAIL).
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
We knew by this time that perception was difficult, and that neat mapping from perception to certainty was impossible.
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
Forty years of AI research applied to factory robots has had almost no practical impact.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
But I am pretty confident that the things that will date my predictions are not yet known by most researchers, and certainly are not the…
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
AI-based robotics systems from research are not used at all in industry.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
We’re still a long way off.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
Traditional robotics trades on geometric relationships and coordinates, which means they are easy to compose, but they are very poor in…
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
Behavior-based robotics has led to more than 20 million robots in people's homes, more than any other kind of robot ever built.
Factual · 15 Jul 2018 |
Technology |
| Not checkable |
Not checkable |
15 Jul 2018 |
Pattern recognition and learning have today clearly moved beyond search.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
I recommend the AI Magazine, the quarterly publication of the Association for the Advancement of Artificial Intelligence.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
But neither of us had very good statistical models, and mine was definitely worse than Raja’s.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
And really none of these groups have any idea how to get there.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
I am perhaps lacking a little humility in claiming this as one of the four approaches to AI.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
we would be extraordinarily surprised to see a robot dolphin emerge from the ocean, one that had been completely designed and constructed…
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
I think it is fair to say that those end to end goals have gotten a little lost over the years.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
And this is the critical problem with symbolic Artificial Intelligence, how the symbols that it uses are grounded in the real world.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
Often, then, these sorts of AI systems have had to deliver much less reliability than autonomous systems will demand in the future.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
No matter how far from reality the sentience of behavior-based systems may be, it is the best we have got.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
That this sort of training works as well as it does is really a little fantastical.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
This ability to be part of the world, and to have agency within it is at some level an artificial sentience.
Not checkable · 15 Jul 2018 |
|
| Mostly true |
Factual |
15 Jul 2018 |
Kim's research team made their legged robot completely blind to show what its low-level reactive behaviours could do.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
Human level intelligence may be just as, or even more, challenging.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
But I also think that people confused performance with competence
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
And neural networks simply suck (yes, suck) at spatial understanding.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
I suspect that human level AI falls into this class.
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
Behavior trees are used underneath two thirds of the world's video games.
Factual · 15 Jul 2018 |
Technology |
| Not researched |
Factual |
15 Jul 2018 |
Artificial neural networks are loosely based on a circa 1948 understanding of neurons in the brain.
Factual · 15 Jul 2018 |
Science and research methods |
| Mostly true |
Factual |
15 Jul 2018 |
Gravitational waves, predicted by Einstein in 1916, were first detected in 2015, ninety-nine years later.
Factual · 15 Jul 2018 |
Science and research methods |
| Not researched |
Factual |
15 Jul 2018 |
The deep learning network that produced the Frisbee photo caption cannot answer questions about the scene, such as whether the weather suits playing Frisbee.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
Perhaps my prescriptions for research towards Super Intelligence will also turn out to be wrong before very long.
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
In deep learning networks, a human designer chooses the number of layers, how layers connect, and the network's inputs.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
This is why today we are starting to think it natural to be able to talk to our machines.
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
SLAM algorithms became part of the basis for self-driving cars, and subsystems derived from SLAM are used in all such systems.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
As usual, I think I may have grossly overestimated the current capabilities of AI systems.
Not checkable · 15 Jul 2018 |
|
| Resolved true |
Factual |
15 Jul 2018 |
In 1977 Rodney Brooks joined the Hand-Eye group at the Stanford AI Laboratory, working on vision for his PhD.
Factual · 15 Jul 2018 |
Their own work or organisation |
| Not researched |
Factual |
15 Jul 2018 |
Learning in such neural networks consists of selecting hundreds of thousands, perhaps millions, of numerical weights.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
These descriptions are meant to give just a flavor.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
I think this is when people really started to take notice of Deep Learning.
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
15 Jul 2018 |
Note that under this evaluation a human child scores six hundred points whereas the four AI approaches score a total of eight or nine…
Not checkable · 15 Jul 2018 |
|
| Not researched |
Factual |
15 Jul 2018 |
Soon after the 2006 Deep Learning revolution, there were practical neural networks with twelve layers.
Factual · 15 Jul 2018 |
AI capabilities |
| Not checkable |
Not checkable |
15 Jul 2018 |
We have been actively trying for 62 years, but apparently it is only just right now that we are about to make all the breakthroughs that…
Not checkable · 15 Jul 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
The true test of the viability of driverless cars will be when they are not just in testing or in demonstration, but when the owners of…
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
Turning them into reality is hard.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
I think the popular interpretation is that driverless cars will simply replace cars with human drivers, one for one.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
I will need a whole new blog post to explain this
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
I am going to take responsibility for what I say, and make it so that people can hold me to whether I turn out to be right or wrong.
Not checkable · 1 Jan 2018 |
|
| Resolved true |
Factual |
1 Jan 2018 |
Academic rumblings about the limits of Deep Learning are already happening as of 2018.
Factual · 1 Jan 2018 |
AI capabilities |
| Not yet due |
Prediction |
1 Jan 2018 |
There is a real possibility that a flying car will not be purchasable by any US resident with enough money even by 2050 (original prediction: not before 2036).
Prediction · 1 Jan 2018 |
Technology |
| Not checkable |
Not checkable |
1 Jan 2018 |
It does however build on large bodies of prior work, and therefore it is more likely to succeed.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
Sorry, it is just going to take longer than every one expects.
Not checkable · 1 Jan 2018 |
|
| Not yet due |
Prediction |
1 Jan 2018 |
Hyperloop will not happen in Brooks's lifetime, i.e. not by 2050.
Prediction · 1 Jan 2018 |
Technology |
| Not researched |
Factual |
1 Jan 2018 |
Brooks said that, despite some impressive lab demonstrations, we have not actually seen any improvement in widely deployed robotic hands or end effectors in the last 40 years.
Factual · 1 Jan 2018 |
Technology |
| Unresolvable |
Prediction |
1 Jan 2018 |
Point-to-point transport on Earth in about an hour using a BF rocket will not happen in the author's lifetime without a major new breakthrough.
Prediction · 1 Jan 2018 |
AI timelines and forecasting |
| Not checkable |
Not checkable |
1 Jan 2018 |
For completely new ideas, however, it is much harder to predict with confidence that the technologies will become deployable in any…
Not checkable · 1 Jan 2018 |
|
| Not researched |
Factual |
1 Jan 2018 |
Machine learning was not used to get to the Moon in the Apollo era.
Factual · 1 Jan 2018 |
Technology |
| Not checkable |
Not checkable |
1 Jan 2018 |
Building human level intelligence and human level physical capability is really, really hard.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
In reality we are less than 1% of the way there, with no real intellectual ideas yet on how to get to 5%.
Not checkable · 1 Jan 2018 |
|
| Not yet due |
Prediction |
1 Jan 2018 |
The generally agreed 'next big thing' in AI beyond deep learning will be something already being worked on with published papers, emerging no earlier than 2023 and by 2027.
Prediction · 1 Jan 2018 |
AI capabilities |
| Unresolvable |
Prediction |
1 Jan 2018 |
Realising the hyperloop will require many new technologies and designs, none of which existed before, for every aspect of the system.
Prediction · 1 Jan 2018 |
Technology |
| Not checkable |
Not checkable |
1 Jan 2018 |
But adoption is still a ways off.
Not checkable · 1 Jan 2018 |
|
| Unresolvable |
Prediction |
1 Jan 2018 |
Underground car-pallet transport at over 100mph will not become a real, viable mass-market service in the author's lifetime, only small demonstration projects.
Prediction · 1 Jan 2018 |
Technology |
| Not checkable |
Not checkable |
1 Jan 2018 |
I do not think that I am a techno-pessimist.
Not checkable · 1 Jan 2018 |
|
| Unresolvable |
Prediction |
1 Jan 2018 |
Cities will be changed by driverless cars through special lanes, geo-fencing, altered pick-up/drop-off norms, parking regulation changes and other incremental modifications.
Prediction · 1 Jan 2018 |
Technology |
| Not researched |
Factual |
1 Jan 2018 |
There has been no improvement in widely deployed robotic hands or end effectors in the last 40 years (as of 2018).
Factual · 1 Jan 2018 |
AI capabilities |
| Not checkable |
Not checkable |
1 Jan 2018 |
Rather, I think of myself as a techno-realist.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
It always takes longer than you think. It just does.
Not checkable · 1 Jan 2018 |
|
| Not yet due |
Promise |
1 Jan 2018 |
Rodney Brooks will publish on his blog dated predictions covering the next thirty-two years.
Promise · 1 Jan 2018 |
Their own work or organisation |
| Resolved true |
Factual |
1 Jan 2018 |
As of the end of 2017, IPv6 adoption was far from complete, with network traffic categories on IPv6 ranging from under 2% to just over 20%.
Factual · 1 Jan 2018 |
Technology |
| Not yet due |
Prediction |
1 Jan 2018 |
No self driving car will be faced with an instance of the trolley problem until beyond 2050.
Prediction · 1 Jan 2018 |
Philosophy, ethics and society |
| Not checkable |
Not checkable |
1 Jan 2018 |
But soon the houses of the elderly will be cluttered with too many robots.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
It will be magical for the human race if this happens by then.
Not checkable · 1 Jan 2018 |
|
| Not yet due |
Prediction |
1 Jan 2018 |
Flying cars will reach 0.01% of total US cars no earlier than 2042.
Prediction · 1 Jan 2018 |
Technology |
| Not researched |
Factual |
1 Jan 2018 |
Rodney Brooks's 1977 Master's thesis used Markov chains to prove the convergence of a particular machine learning algorithm.
Factual · 1 Jan 2018 |
Their own work or organisation |
| Not checkable |
Not checkable |
1 Jan 2018 |
Unless parking and human drivers are banned from those areas before then.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
A common thread that I have observed is how people tend to underestimate how long new technologies will take to be adopted after proof of…
Not checkable · 1 Jan 2018 |
|
| Resolved true |
Prediction |
1 Jan 2018 |
Claims before 2023 to be the next big thing in AI beyond deep learning will not pan out.
Prediction · 1 Jan 2018 |
AI capabilities |
| Not yet due |
Prediction |
1 Jan 2018 |
There will be no commercially viable passenger-carrying hyperloop systems within the next 32 years (by 2050).
Prediction · 1 Jan 2018 |
Technology |
| Not researched |
Factual |
1 Jan 2018 |
No technical obstacles stood in the way of adopting IPv6.
Factual · 1 Jan 2018 |
Technology |
| Not yet due |
Prediction |
1 Jan 2018 |
Rodney Brooks said he thinks landing cargo on Mars by 2026 is optimistic, though it might be pushed to happen as a statement that it can be done rather than for a pressing practical reason.
Prediction · 1 Jan 2018 |
AI timelines and forecasting |
| Not checkable |
Not checkable |
1 Jan 2018 |
It was an abysmally terrible thesis.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
This is so much harder than most people imagine it to be--many think we are already there; I say we are not at all there.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
I would have been wildly over optimistic about it.
Not checkable · 1 Jan 2018 |
|
| Not researched |
Factual |
1 Jan 2018 |
As of 2018, Uber, Lyft and conventional taxi services offer taxi service with arbitrary pick-up and drop-off locations in major US cities.
Factual · 1 Jan 2018 |
Economics and markets |
| Mostly true |
Factual |
1 Jan 2018 |
Horseless carriages did not simply replace horse-drawn carriages one for one but required new infrastructure and ownership and usage models.
Factual · 1 Jan 2018 |
Technology |
| Not checkable |
Not checkable |
1 Jan 2018 |
For me it seems clear that driverless cars are not going to simply be the same sorts of cars as normal cars, but simply without human…
Not checkable · 1 Jan 2018 |
|
| Unresolvable |
Prediction |
1 Jan 2018 |
Robotics, AI and ML will not produce a general intelligence that can suddenly do all the things humans or chimpanzees can do.
Prediction · 1 Jan 2018 |
AI capabilities |
| Not checkable |
Not checkable |
1 Jan 2018 |
Building flying cars, or a hyperloop system (or a palletized underground car transport network underground) is hard.
Not checkable · 1 Jan 2018 |
|
| Resolved true |
Factual |
1 Jan 2018 |
Rodney Brooks previously identified underestimating adoption time of new technologies as the seventh of seven deadly sins of predicting the future of AI.
Factual · 1 Jan 2018 |
Their own work or organisation |
| Not checkable |
Not checkable |
1 Jan 2018 |
I may well have inflated them by a factor of 10 or more, and for that I apologize.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
There has been a little tiny burst of progress over the last five years, and too many people think it is all done.
Not checkable · 1 Jan 2018 |
|
| Unresolvable |
Prediction |
1 Jan 2018 |
Driverless cars will not simply replace human-driven cars one for one.
Prediction · 1 Jan 2018 |
Technology |
| Not yet due |
Prediction |
1 Jan 2018 |
A flying car will not be available for purchase by any US resident with enough money before 2036, and there is a real possibility this will not happen at all by 2050.
Prediction · 1 Jan 2018 |
Technology |
| Not researched |
Factual |
1 Jan 2018 |
Brooks says he was right that the next big thing in AI might show up in 2023, and that papers about it already existed before 2018.
Factual · 1 Jan 2018 |
AI capabilities |
| Not checkable |
Not checkable |
1 Jan 2018 |
It does not save anything for later to be used for a similar problem at some future time.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
If something has not even been demonstrated yet in the lab, even though the physics says that it will be good to go, then I think it is a…
Not checkable · 1 Jan 2018 |
|
| Resolved true |
Factual |
1 Jan 2018 |
In the months before January 2018, Rodney Brooks publicly criticised what he considers hype around AI and machine learning.
Factual · 1 Jan 2018 |
Their own work or organisation |
| Not checkable |
Not checkable |
1 Jan 2018 |
The geographies will have to be where the roads are wide enough for other drivers to get around stopped vehicles.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
Building electric cars and reusable rockets is easy.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
I would have thought that IPv6 would have been universal by 2010 or so.
Not checkable · 1 Jan 2018 |
|
| Not checkable |
Not checkable |
1 Jan 2018 |
Those of you who have been reading my series of blog posts on the future of Robotics and Artificial Intelligence know that I am more…
Not checkable · 1 Jan 2018 |
|