A 16 year old has been training for almost 16 years to drive a car. I would argue the opposite: Waymo’s / Specific AIs need far less data than humans. Humans can generalize their training, but they definitely need a LOT of training!
When humans, or dogs or cats for that matter, react to novel situations they encounter, when they appear to generalize or synthesize prior diverse experience into a novel reaction, that new experience and new reaction feeds directly back into their mental model and alters it on the fly. It doesn't just tack on a new memory. New experience and new information back-propagates constantly adjusting the weights and meanings of prior memories. This is a more multi-dimensional alteration than simply re-training a model to come up with a new right answer... it also exposes to the human mental model all the potential flaws in all the previous answers which may have been sufficiently correct before.
This is why, for example, a 30 year old can lose control of a car on an icy road and then suddenly, in the span of half a second before crashing, remember a time they intentionally drifted a car on the street when they were 16 and reflect on how stupid they were. In the human or animal mental model, all events are recalled by other things, and all are constantly adapting, even adapting past things.
The tokens we take in and process are not words, nor spatial artifacts. We read a whole model as a token, and our output is a vector of weighted models that we somewhat trust and somewhat discard. Meeting a new person, you will compare all their apparent models to the ones you know: Facial models, audio models, language models, political models. You ingest their vector of models as tokens and attempt to compare them to your own existing ones, while updating yours at the same time. Only once our thoughts have arranged those competing models we hold in some kind of hierarchy do we poll those models for which ones are appropriate to synthesize words or actions from.
I meant visual patterns, too. You're thinking about what I said on too granular a level. JEPA is visual, based ultimately on pixels. The tokens may be digested from pixels until they're as large as whole recognizable objects, but the tokens are not whole mental models themselves.
Here's an example of humans evaluating competing mental models as tokens: You see a car, it's white, it's got some blood stains on the door, and it's traveling towards a red light at 90 miles an hour in a 30 mph residential zone, while you're about to make a left turn. A human foot is dangling from the trunk.
You refer to several mental models you have about high speed chases, drug cartels in the area, murders, etc. You compare these models to determine the next action the car might take.
What were the tokens in this scenario? The color of the car, the pixels of blood, the speed, the traffic pattern? Or whole models of understanding behavior where you had to choose between a normal driver's behavior and that of someone with a dead body fleeing a crime scene?
They were practicing object recognition, movement tracking and prediction, self-localisation, visual odometry fused with porpiroception and the vestibular system, and movement controls for 16 years before they even sit behind a steering wheel though.
That's an exaggeration. Nobody is trained to read STOP signs for 16 years, a few months top. And Waymo doesn't need to coordinate a four-limbed, 20-digited, one-headed body to operate a car.
Well, I also think that there is a lot that we process 'in background' and learn on beforehand in order to learn how to drive and then drive. I think the most 'fair' would be to figure out absolute lowest age of kids that would allow them to perform well on streets behind steering wheel.
i am not making a point that it is, I am rather expanding on the possible perspective in which 16 years of training produce a human driver.
That being said, you don't really need training to understand a STOP sign by the time you are required to, its pretty damn clear, it being one of the simpler signs.
But you do get a lot of "cultural training" so to speak.
Difference is pretty big if it’s icy like breaking 100 meters vs 10 meters. Especially if there’s wildlife like reindeers/moose’s you are going to do emergency breathing semi regularly.
Just pointing out - a lot of snowy areas are very aggressive about plowing (and salting). For most people this is probably like "don't drive tomorrow" and not some need for knobby snow tires.
Even when the road is dry the rubber compound is a lot softer on winter tires so you get significantly more grip than all season or summer in cold temps when they get hard.
There are laws of physics you can’t hand waive away. Winter tires are really more cold temp tires. The rubber formulation is different to allow for grip in the cold and dry (tread pattern for cold and snow). As such a winter tire wears heavily driven in the summer, rubber formulation is just too soft.
For an all season that level of summer wear would be unacceptable. So a different formulation is used to improve summer wear at the cost of the winter low temp performance. You can’t have it both ways, a long wearing summer performance and good sub 40 degree grip.
Modern high quality all weather tyres are excellent in summer and winter.
Except on actual snow, where they're just ok, because of the hybrid sipe patterns, and ice, where they suck exactly as much as everything else except studded tyres (which suck on tarmac instead).
Even after the streets are plowed there's still a bit of snow on the main roads and a lot on the side roads. Maybe you live in a place with really mild winters, but my car would have drifted into a ditch many times this winter if I didn't swap my tires.
I... Well, I had started explaining point by point how wrong this is but frankly the answer is just "all of it, very".
I've driven summer tires, all season tires, winter tires, and studded winter tires in every season in Canada. (Yes, I live in Canada and own borderline-usless summer-only tires. Yes, I've tried driving them in snow.)
None of what you're saying lines up with my own experience, various YouTube videos on braking distances, or literally anything else I've ever seen anywhere.
Edit: And, well, to be clear... I've lived on the West coast of Canada where it's a bit more mild but you're in the mountains, in the middle where it hits -50, and in the East where it only hits -30 but snows like hell.
Yes, there are. And they show that it's a trade-off well worth investigating. Do you really want 10% better performance on snow at the cost of 10% worse performance on tarmac?
How much do you drive on snow anyway? Probably nowhere near as long as you do on tarmac, even in a tough winter.
It’s more about operational resilience and serving customers than product development. If you run early WhatsApp like organisation just 1 person leaving can create awful problems. Same for serving customers especially big clients need all kinds of reports and resources that skeleton organisation can not provide.
Yeah, that’s a misconception too based on my experience.
I’ve seen many people (even myself) thinking the same: if I quit/something happens to me, there will be no one who knows how this works/how to do this. Turned out the businesses always survived. There was a tiny inconvenience, but other than that: nothing. There is always someone who is willing to pick up/take over the task in zero amount of time.
I mean I agree with you, in theory. But that’s not what I’ve seen in practice.
Utter bullshit engineered to convince students not to do drugs. Adderall doesn't make you magically better at solving the knapsack problem, it's not NZT-48 from Limitless. That's not why anyone takes it.
The two most helpful things with SQL are (1) always use set-based operations (never a cursor) and (2) break up your queries into smallest possible reusable chunks (CTEs). Sprinkle in tests to taste. Without some discipline SQL can get out of hand. This is what made dbt popular.
Can you share more detail? Every pain point I've seen can be handled by a more suitable/performant/flexible DB and/or software that makes working with SQL less painful.
reply