This is great, but what is a possible use-case of these massive classifier models? I'm guessing they won't be running at the edge, which precludes them from real-time applications like self-driving cars, smartphones, or military. So then what? Facial recognition for police/governments or targeted advertisement based on your Instagram/Google photos? I'm genuinely curious.
I imagine this is a stage in a larger plan to incorporate an AI assistant to excel? Being able to talk to your spreadsheet in natural language and have it do a bunch of analysis and visualizations would be a huge productivity booster.
That's the question I'm asking myself too. I live in Seattle, where you can't walk two blocks without seeing a new (wooden) residential building going up. I'm sure that newer buildings like these take more precautions than the historic ones in Lahaina, but it still makes me wonder, since Seattle isn't historically prone to wildfires.
If I remember correctly, the algorithm in Alpha Go used a combination of reinforcement learning and searching the space of moves and possible outcomes to determine the next move. So it does indeed use both search and learning.
Regardless, the author's point is that computation is a better way of finding and exploiting patterns/strategies than our own intuitions. The distinction between search and learning is not the important one here.
There was an important step prior to alpha go. At the time the combinatorics were in favor of Go. But someone had the bright idea to do a probabalistic search of the space. The key idea was to play a ton of random games and rate each position based what percentage that spot was included in winning games. This blew away all other go ai at the time. Sadly this was about the time I stopped having time to follow the space, so I’m not sure how this idea was further incorporated in Go AI. But it was truly a revolutionary idea at the time
In hindsight computation wasn't the important thing though. A lot of things require a lot of computation that aren't intelligent or don't scale well, like Deep Blue. The important breakthrough in AI was learning ("machine learning").
Search/Reasoning/Inference time compute, however you phrase it is still essential. You need search to improve upon learning to work in novel situations.
I imagine that carbon footprint could be reduced substantially by this too, at least on a per-watt of utilized energy basis. Imagine if all computer chips used semiconductor materials - then more of the input electricity is actually put towards computing, and cooling fans are a thing of the past!
I've only tried a few songs but they've mostly been bangers! I did come across a couple examples where the recommended songs just heavily sampled the original but overall very impressed.
I think this line from the article serves as a good summary:
> We’ve never seen anything like this in California housing history where a residential building of any height, with any amount of parking, can be placed in the wealthiest communities in the world provided its just 20% affordable and is safe.
In such a safety-critical system, it's important to have backups. You need at least two cameras to accurately estimate depth, and cameras can fail for a variety of reasons (sun glare, low lighting, heavy fog). Radar, at the very least, is a backup for the cameras. Also, with vision, the best you can do is estimate distance, whereas with radar and LIDAR you are explicitly measuring it.
>Also, with vision, the best you can do is estimate distance, whereas with radar and LIDAR you are explicitly measuring it.
But is there any evidence you need to measure distances? We humans can navigate the world without walking into walls, so long as we're looking where we're going. For a machine to navigate the world it should be possible to do it via vision. And Tesla's do have multiple cameras to be able to measure depth.
And radar is not a backup for cameras. The resolution of the data is terrible and you can not rely on it to do any sort of driving except braking if it thinks there's an obstacle. Radar is also susceptible to problems as well, which is why Tesla's and other cars with radar can often go crazy thinking you're gonna crash randomly.
Not saying all of that is true. It seems to be pretty much philosohpy at this point. But maybe one day our methods scale up enough or evolve :-) in some other way and there will be a breaktrough in evolutionary AI, like there was in neural nets.
