Hacker News Reader

Comment by dvtrn | Hacker News Reader

> The dog has a lot of false positives from the cameras being triggered by car headlights or small animals.

My real dog has the same feature, so you’re at least doing a good job mimicking nature, heh.

Haha came here to say the same thing. My stupid dog barks at everything. The wind, the sun, the rain, ants, if he thinks im home, if he thinks im leaving, if he sees me, if he doesnt see me, he even barks in his friggin sleep. Does he bark when he needs to take a piss or a shit and needs to go outside? Nope. Prick.

Train your dog to piss or shit. If your dog likes food it should be very possible.

The dog also doesn't come with an off switch in case you have an important meeting with your CEO, taking it in your home office and there's a squirrel running around...

Oh dear. Can’t say I’ve had that problem...really ever.

How many squirrels have you had to chase/escort out so far?

Oh, I'm sure a squirrel outside is plenty to cause a problem. My dogs have historically not needed anything at all to actually be moving outside to trigger an endless barking spree.

The barking sure (though I set up my office in the basement because FAMILY LOUD and so when they go off, it's pretty well muffled and people on zoom calls don't even hear it), it was the "bringing a squirrel inside and then setting it loose in the house" part that I was wondering about, ha.

I think you're confusing some awkward phrasing from the poster. They were taking the meeting with the CEO in their home office, and the squirrel was running around presumably outside causing the dog to bark.

But I can see how you read it as thet were having a meeting and the dog brought a squirrel into the home office and ran around.

That would be worse, but if we let the dog actually catch the squirrel it won't be running anymore... But we try very hard for that not to happen.

No I meant it is running around outside. Often literally around because there are fruit trees in the garden around the exit from the house, and there's a big glass door where the dog can see outside. The squirrels learn pretty quickly that we don't want the dog to actually catch the squirrel (who knows what it has eaten and what diseases it may have), but to the dog it's hard to see squirrel just going about its business right under his nose...

Yep. Especially given big glass doors and squirrels that quickly realize barking dog behind the glass door poses no danger whatsoever, so it's ok to proceed with their business.

One of the cats in our neighborhood likes to sit in front of our glass doors to antagonize our dog. The dog goes nuts and the cat sits there washing itself.

I don't know about during a meeting, but my count of squirrels I've had to chase/escort out of my house currently stands at two. Although only one escort was successful. The other one I thought had left, but I found its desiccated body under the ashes in my fireplace about a week later.

Certainly a feature. I recommend a sensitivity level which uses breeds for each level. I further recommend the highest sensitivity be "Great Pyrenees" as I can attest my dog barks at every false positive.

Came here to say that :) Our recently-adopted rescue terrier will bark at anything. Right now he's especially targeting birds that dare to land on our (his!) bird feeder.

It's definitely a feature. Next door get a delivery? Trigger!

Interesting! They might be both working from a similar algorithm.

That reminds me of the argument about sentience lately. Why can't a neural network be sentient if all our actions have been trained by interacting with the world since birth?

If our minds are like computer neural networks, then we can only achieve sentience by priming an existing "human person" neural model with data, which is how every human has worked.

The operative word in your comment being "actions". What "actions" can (current generation) neural networks undertake? Can they initiate anything? Can they choose their own learning material? Can they even choose when or whether to repeat a certain training set?

There was a recent HN comment about this that I think illustrates the point well [0]:

experimentation is an act on the world to set its state and then measure it. That's what learning involves. These machines do not act on the world, they just capture correlations.

[0] https://news.ycombinator.com/item?id=32201757

Wouldn't you say that a neural network that learns to play a game by playing the game does so via experimentation?

A lot of human "choice" is actually just a neuron firing or chemical in the brain going off

Because most people define sentience as 'a black box that reacts meaningfully to inputs in a sufficiently unpredictable and inexplicable manner', and for most people (especially researchers in these fields), current neural networks do not meet this bar.

Reply on news.ycombinator.com