dwaltrip

33 ms·

661.
by dwaltrip 3y ago
It’s not about groveling. It’s about demonstrating that you have the skills and experience they are looking for. You are selling something, why should anyone buy it? Convince them.
662.
by dwaltrip 3y ago
> we're slowly entering the "tooling" phase of AI, where people realize there's no real value generation here Hard disagree. A very clear counterexample from my usage: Gpt-4 is phenomenal at helping a skilled person w
663.
by dwaltrip 3y ago
This such a good example of why communication is hard. You’ve made the distinction pretty clear, but often times it isn’t so easy to see or it can get muddled.
664.
by dwaltrip 3y ago
Bard seems worse than gpt-3.5 in my experience.
665.
by dwaltrip 3y ago
For those in the know, any comments about the team? Looks impressive to my uniformed eyes.
666.
by dwaltrip 3y ago
Copying itself means it can do more “work” in the same amount of time. Earning money, strategizing, researching, smooth-talking humans, playing politics, and so on. Potentially even improving itself. These activities could give it power and
667.
by dwaltrip 3y ago
On your specific point, humans take decades to (very roughly) copy themselves. An AI could do it in minutes with far more precision and control. Assuming you have 200 IQ, if you could make thousands or millions of clones of yourself quite c
668.
by dwaltrip 3y ago
Thanks for double checking. I ran out of steam on the convo heh.
669.
by dwaltrip 3y ago
If someone close to you was getting so obsessed with longevity that it was negatively affecting the other aspects of their life, then it might be reasonable to talk them about getting professional help. It could be a difficult discussion, b
670.
by dwaltrip 3y ago
That’s super interesting. I’m recovering from burnout and other health issues and I’ve found it to be occasional helpful in the way you are describing. For me, it can smooth out the process and “lower the intensity” of accomplishing any par
671.
by dwaltrip 3y ago
Do you think your comments are effective? People don’t generally react well to someone telling them “your idea is stupid, you need to see a therapist”.
672.
by dwaltrip 3y ago
Would multiple bars meaningfully degrade the performance of the jet?
673.
by dwaltrip 3y ago
The last paragraph in the article: > That such big performance increases can be extracted from relatively simple changes like rounding numbers or switching programming languages might seem surprising. But it reflects the breakneck speed
674.
by dwaltrip 3y ago
Ah, oops, that should say 2018, not 2023.
675.
by dwaltrip 3y ago
The linear plot in that Wolfram link is messed up. It doesn't show all the data (caps out at 800 billion GDP). Here's a corrected linear plot, from the script that I linked (commenting out the log-log scaling): https://
676.
by dwaltrip 3y ago
I started using Spotify for podcasts occasionally, and it hasn’t seemed to clutter anything up for me. But if you already have a podcast app you love, that makes sense.
677.
by dwaltrip 3y ago
The countries with the 2 highest GDPs, China and the US, are in the top 4 largest countries. India, #5 in GDP, is #7 in area. I guarantee that if you plot the countries of the world by GDP and area, you will see a trend line. It also makes
678.
by dwaltrip 3y ago
I just learned about the equal earth projection, which was apparently invented very recently in 2023. It’s the best looking equal area projection I’ve seen. https://en.m.wikipedia.org/wiki/Equal_Earth_projection
679.
by dwaltrip 3y ago
My experience has differed from yours then. I think you overstate the frequency of which it hallucinates for simpler questions. I am usually using GPT-4, which is more reliable. For the types of questions I’m talking about, I’d say GPT-4 is
680.
by dwaltrip 3y ago
Its trustworthiness is not near zero… There are many categories and types of questions where you can have a good degree of confidence in the answer. This is especially true for questions where a rough approximation is acceptable. Off the to
681.
by dwaltrip 3y ago
Is that real? Hilarious if so.
682.
by dwaltrip 3y ago
First comment in 13 years! Amazing. Nice to have ya around. Enjoy your lurking :)
683.
by dwaltrip 3y ago
I’ve heard the idea that carbon credits create a fairly unique scenario where both sides, the buyer and the seller of the credits, aren’t that strongly incentivized to call out fraud. In regular “marketplace” transactions, the buyer wants t
684.
by dwaltrip 3y ago
Hey, I wanted to reply to this, but just didn't find the time. I disagree with a lot of what you wrote, but your arguments are stated clearly. I appreciate that. Cheers :)
685.
by dwaltrip 3y ago
Look at the interesting discussion that happened as a result of my comment. My comment was completely fine. Your incredulity and exasperation is misplaced. If you can’t realize that then I think we are done here. Btw, since we are already i
686.
by dwaltrip 3y ago
Can you please argue your case? I argue that the model’s weights could contain a rich representation of the world, even if it looks different from that of humans. And that it can use this capability of world modeling to also make new conne
687.
by dwaltrip 3y ago
Can you propose more tests of “understanding”? The more concrete the better. Or is more like “you know it when you see it”? I tested GPT-4 with your example of sky color / human eyes. It performed quite well and seemed to have a pretty
688.
by dwaltrip 3y ago
That makes sense, thanks for the detailed reply. I’m curious about the latest research related to continuously improving the tokenizer while training the main model (or something equivalent).
689.
by dwaltrip 3y ago
What does it mean to actually understand why the sky is blue? What are the specifics characteristics of a hypothetical AI system that you would feel comfortable giving the label of “understanding” to?
690.
by dwaltrip 3y ago
Hmmm, is there any way for a model to predict / edit its own tokens as it learns? Kind of a recursive, self-evolving tokenization process? I’m pretty new to deep learning, so this analogy might be off. But I’m reminded of convolutional

More ›