jebarker

26 ms·

511.
by jebarker 2y ago
Having a kid may have dinged my raw intelligence but the forced ruthless prioritization of what I do and what I value doing has actually made me more competent at home and work
512.
by jebarker 2y ago
I also think we've become more self-obsessed and the idea that we might need to sacrifice time spent on ourselves scares us.
513.
by jebarker 2y ago
I'm sorry, you're correct, I missed that. I'll edit my edit!
514.
by jebarker 2y ago
I'd like this too, especially if it generated binaural audio.
515.
by jebarker 2y ago
That would be explained by data leakage too, e.g. sampling frames in the train and test data from the same video sequences. There's nothing in the writeup that says suggests the model is explicitly matching audio to ground truth images
516.
by jebarker 2y ago
I think you are misunderstanding. I don't think the network matches the audio to a ground truth image and generates an image. It just takes in audio and predicts an image. They just use the ground truth images for training the model an
517.
by jebarker 2y ago
I'm a little surprised by the seeming enthusiasm in the report for TTT as an approach. The results speak for themselves and TTT seems like a powerful approach. But the dependence on large amounts of synthetic pre-training data seems to
518.
by jebarker 2y ago
That's a good point. But re: not anthropomorphizing, what's wrong with errors, mistakes or inaccuracies? That's something everybody is familiar with and is more accurate. I'd guess most people have never actually experie
519.
by jebarker 2y ago
I agree with your point, but I don't think anthropomorphizing LLMs is helpful. They're statistical estimators trained by curve fitting. All generations are equally valid for the training data, objective and architecture. To me it&
520.
by jebarker 2y ago
I see, thank-you for clarifying.
521.
by jebarker 2y ago
I guess I still don't understand though. Is the claim that tech innovation in the rest of the world is being turned into money in the US only? If so, why is that happening?
522.
by jebarker 2y ago
Can you expand on this? What do you mean by "monetized in the US stock market"?
523.
by jebarker 2y ago
Really? The Dreambooth results don't seem to have really changed the input image. The age change from the paper looks better to me but not accurate to reality.
524.
by jebarker 2y ago
Naive question here: wouldn't the upside of keeping this secret and trading on it's output yourself far exceed what you can make in subscriptions?
525.
by jebarker 2y ago
Do you have any evidence to support that claim? Competitive programmers (especially those with their own libraries ready to go) can be incredibly fast at solving coding challenges.
526.
by jebarker 2y ago
I always get too ambitious with AoC, e.g. trying to do it in a new to me language or abstract everything into a new library. Subsequently I never finish. So this year I'm using whatever language or tool is easiest for me and the only g
527.
by jebarker 2y ago
I highly recommend watching him speak and debate too.
528.
by jebarker 2y ago
I believe this will happen in many human domains, but it doesn't really matter. Nobody is going to stop writing poetry because of this and I doubt there's much of an audience for AI generated poetry.
529.
by jebarker 2y ago
I wonder how well experts in poetry would do? It seems like a strange area to use non-expert humans for evaluation because these days most people reading any poetry are at least enthusiasts.
530.
by jebarker 2y ago
> A simple way to do this is to make playlists for a given key, which helps reinforce the sense of shared tonality across songs. This is gold. I similarly maintain lists of melodies that prominently feature certain intervals.
531.
by jebarker 2y ago
Why isn't LLM pre-training based on next token prediction considered "trial and error"? It seems to fit that description pretty well to me.
532.
by jebarker 2y ago
> In essence, it's like a neural net learning to follow a recipe by watching a chef cook—rather than inventing its own recipes entirely from first principles. Just like how a chef learns
533.
by jebarker 2y ago
I'm guessing lots of people would argue Google doesn't really get credit for that.
534.
by jebarker 2y ago
I agree that most coders would be well served to learn some linear algebra and probability theory as those are ubiquitous. Calculus and stats are also useful depending on your specialty. LLMs don't really change any of this.
535.
by jebarker 2y ago
It's hard to know the right questions to ask to explore these reasoning models. It's common for me to ask a question that's too easy or too hard in non-obvious ways.
536.
by jebarker 2y ago
As far as I'm aware the "self-service" version of it on their website is in the past year. I think you're right that it's been possible over the phone for a long time though.
537.
by jebarker 2y ago
The best feature Comcast has added in recent years is the ability to request a $5 bill credit each time there's an outage. I use it more months than not. I suppose that's good customer service although it'd be better if it wa
538.
by jebarker 2y ago
With an inReach I have the option of periodically tracking my position and uploading that to a site my loved ones can check. Even whilst doing this I can leave the device on for a multi-day trip without worrying about battery drain. I'
539.
by jebarker 2y ago
Initially I thought the same re: hiking, skiing etc. The only issue I see is that cellphone battery life is terrible compared to inReach like devices. Not sure I'd want to depend on it for longer than a few hours.
540.
by jebarker 2y ago
How suggestive is this of life elsewhere in the universe?

More ›