resters

35 ms·

511.
by resters 3y ago
The biggest factor in whether people become rich is simply luck. Why else would a small number of people who are by far not the most intelligent or capable among us, and who happen to live in a few select countries, be the ones who own 90+%
512.
by resters 3y ago
Point taken. Maybe LLM beings would be the driving force behind more widespread adoption of ECC RAM.
513.
by resters 3y ago
> we could embody an AI ... a hardware interface that generates a token stream from a living human's body would seem to enable this at some level. Not sure how it would work at scale. Maybe something much simpler like phones with b
514.
by resters 3y ago
Here's my theory: Consider a typical LLM token vector used to train and interact with an LLM. Now imagine that other aspects of being human (sensory input, emotional input, physical body sensation, gut feelings, etc.) could be added as
515.
by resters 3y ago
Not to criticize Sam, but I think people don't realize that it was Greg who was the visionary behind OpenAI. Read his blog. Greg happens to be a chill, low-drama person and surely he recruited Sam because he knew he is a great communic
516.
by resters 3y ago
In my opinion the key takeaway is that compute is becoming commoditized much more rapidly than anyone expected, and that IP is becoming less and less relevant compared to fabrication, energy and land. For consumers of cloud infrastructure,
517.
by resters 3y ago
I thought there were "presets" that NPM and Yarn knew about to do this kind of thing...
518.
by resters 3y ago
Ahh I hadn't seen that detail reported. Thank you.
519.
by resters 3y ago
Won't the pain be suffered primarily by the poorest Argentinians? If transit subsidies are removed, how will people get to their jobs?
520.
by resters 3y ago
> people can't monitor that people aren't breaking the law Typically even without encryption there are laws banning monitoring of such communications. For example, tapping into someone's analog phone line or listening to u
521.
by resters 3y ago
If encryption were allowed, there would be no way to observe that hams were using their licenses in ways that comply with most of the FCC rules that differentiate "amateur" from professional radio use cases. In my view, of course
522.
by resters 3y ago
Maybe there is a way to train a neural network in a distributed way by training subsets of it and then connecting the aggregated weight changes to adjacent network segments. It wouldn't recover 1000x interconnect slowdowns, but might s
523.
by resters 3y ago
The way that official inflation numbers are calculated is a weighted average deemed useful for policymaking. However price increases are inflation, and inflation in one area of the economy can occur simultaneously to deflation in another
524.
by resters 3y ago
I strongly agree, what you are describing is what is needed. I'm also concerned by the chilling effects we've seen lately on speech, starting with attacks on press freedom by government, but also aided by social media firms cooper
525.
by resters 3y ago
One of the best things about Youtube (and now TikTok) is the growing popularity of videos that reveal the shocking extent of police misconduct. Thanks to camera phones we have all become outraged at the way may police treat black people, an
526.
by resters 3y ago
To the extend that systems like chat-GPT are valuable, I expect we'll have open source equivalents to GPT-7 within the next five years. The only "moat" will be training on copyrighted content, and OpenAI is not likely to be a
527.
by resters 3y ago
A similar phenomenon happens with Instacart deliveries. Shoppers will often mysteriously substitute an item with the tiniest one available in the store. Most likely this is to lessen the load associated with delivering the item.
528.
by resters 3y ago
There should be a distinction between doing a migration of a "monolithic" system to "microservices" vs adding functionality to a monolithic system by implementing a "microservice" that will be consumed by the m
529.
by resters 3y ago
True, the choice of slow or fast could dictate aisle/middle/window position, or the boarding groups could be set up as follows: group 1) fast window seaters group 2) fast middle seaters in rows with the window in group 1 group 3)
530.
by resters 3y ago
I think the most sensible way to do it would be to let people self-select into fast boarders and slow borders, let the fast ones get on first and take their seats, then give the slow boarders as much time as they need.
531.
by resters 3y ago
While this is a valid nitpick, I think people sometimes overdo it. Mathematical objects can be both abstractions and systems in and of themselves. They are abstract "enough" to be useful as models, but anyone who infers causality
532.
by resters 3y ago
How is it that we live in 2023 and apple’s native podcast app is pretty mediocre and now Google is killing one that was slightly better.
533.
by resters 3y ago
It's a loss for everyone who had bet against the market in general as well
534.
by resters 3y ago
One thing I find fascinating about GPT-4 (and I'm curious about your take) is that it can not only generate novel, non-trivial code, but it can (upon request) output that code as a base64 encoded string... seemingly all from the model
535.
by resters 3y ago
For anyone who is interested in amateur radio but hasn't had much experience yet, please note that getting some wire high up into a tree is a great way to make HF operation more fun especially when running a low power transmitter.
536.
by resters 3y ago
I can’t believe it could be possible that less data is better than more data when it comes to self driving car sensors. Why not also include audio data and ir/uv data too? I realized today that the Tesla system is noticeably slow to re
537.
by resters 3y ago
I’ve driven a few Teslas lately and have been extremely impressed by the performance and comfort. But the auto steer and fsd preview are definitely underwhelming. It can’t see a red light that is 120’ away to begin slowing down, nor does it
538.
by resters 4y ago
GPT-4 could probably do a pretty decent job of this.
539.
by resters 4y ago
Example please
540.
by resters 4y ago
Revisit this comment in 2 months, 6 months, a year. It’s mainly a matter of when OpenAI allows widespread access to the gpt-4 api and developers have time to build stuff.

More ›