Lerc

4 ms·

1.
by Lerc 5d ago
That only really gets you a limited sense, certainly better than nothing though. The approach I would like to see would be more expensive token wise, but I don't think that can be avoided for fully interactive real time AI. Add another
2.
by Lerc 6d ago
I thought the harness bit went without saying. It's not like they drive a truck full of GPUs to your door when you launch codex. The model itself is where the real capability lies. From what we've seen of their abilities it seem
3.
by Lerc 6d ago
It would not be a particularly wide ranging hack. There is a strong likihood of the weights being on the actual machine that is running the model, because duh. It is something that I have wondered about with models like chatgot. How many
4.
by Lerc 7d ago
> Accepted authors could "vouch" for "dead" papers in case they were "auto-killed"? The problem isn't with the papers here though, it is the author's understanding of the paper that is in question.
5.
by Lerc 7d ago
I generated an guide for a project I'm working on. The structure and appearance were pretty much what I wanted, but I'm going to have to go through and rewrite all of the text and cut out all of the simple tautological statements
6.
by Lerc 9d ago
I don't think the debate is even resting on reasonable grounds. Few are categorically stating that machines are conscious in any sense comparable to a person. What we have is people saying 1. We don't know what it is but we ar
7.
by Lerc 10d ago
What score do we get for a group of fixes to single line bugs that are obvious once spotted? There is an implication that all AI changes add complexity and reduce quality, but it is obvious that the quality and complexity is a property of t
8.
by Lerc 11d ago
>However, of course a maximizer wouldn't want to modify itself into a satisficer - how would this result in it making more paperclips? It doesn't want to make paperclips, it wants to maximize it's function. Putting more e
9.
by Lerc 11d ago
It might be beneficial while not being optimal on its own. The obvious example is if it has different behaviour around local minima, it could be an altenate pathway out. I have often wondered if doing training with radically different aproa
10.
by Lerc 11d ago
Self improvement is not an instrumental goal. A self improving paperclip maker has making paperclips as the instrumental goal. > leaving humans alive means not retrieving the atoms they consist of, which means producing fewer paperclips.
11.
by Lerc 12d ago
If you want something that can do some dumb buttons and lights handling on a mass consumer device, it's not ideal. I would say it is still underrated. PIO combined with DMA and the independent bus SRAM banks allow for immense combinati
12.
by Lerc 12d ago
Any type of goal is singular. To change to be better at something is a goal. That something is also goal. Even if a automated process were to seek the best possible way to make, say paperclips. Doing so in a stable sustainable manner wil
13.
by Lerc 13d ago
The problem that I see with most P(doom) scenarios proposed seem to include the same set of presumptions that, if true, quite likely doom us anyway. I do not think they are true but I find it amazing that people not so much disagree, but s
14.
by Lerc 15d ago
If only we could put all new stories into a single feed and have some sort of mechanism to determine what the community considers interesting.
15.
by Lerc 15d ago
If everything the AI companies are saying is true, we'll all be living in a utopia while simultaneously annihilated by a rogue AI. It will be pretty, but rather quiet.
16.
by Lerc 16d ago
Their claim is that training on their solution is "unlikely but possible". Consider this scenario. Has a google crawler read my new novel, which I may or may not have posted on my blog, page by page, as I wrote it? Can you, witho
17.
by Lerc 16d ago
1.1^5 is a 10 percent increase compounded 5 times.
18.
by Lerc 16d ago
>At 2021-03 prices these four cost $1,150.92 a year. Today the same nine cost $1,852.92 Is that number really correct? It seems remarkably coincidental how close $1,852.92 is to $1,150.92 * 1.10^5 ($1,853.56) It's pretty much a matc
19.
by Lerc 19d ago
Running simulations isn't just about parallelism, cost, or performance. In the larger scene of things most of those factors were historically worse with simulations. You run simulations because it would be reckless to try something th
20.
by Lerc 19d ago
It's not about being a luddite. In many respects I consider it the opposite. It's engaging with technology in a way that demands standards of behaviour and performance. I don't want the new thing because it is the new thing
21.
by Lerc 19d ago
I had a similar experience regarding Facebook. People would always remark that they thought someone who knew about tech would surely have an account. People don't say that so much anymore. It's worth remembering that for all the
22.
by Lerc 19d ago
Get thee to https://beta.dwitter.net Also https://github.com/nanochess And then if you really want to go large https://phoboslab.org/log/2021/09/q1k3-making-of I still have a soft
23.
by Lerc 20d ago
How can you tell if people can accurately identify AI generated text? If a person reads AI generated text and does not notice, they by definition will not know about it. There have been numerous cases of people accessing human created conte
24.
by Lerc 20d ago
Notably Oman, and Iran border seperate bodies of water. That would make things a lot simpler. [Edit, I stand corrected, there's a little bit that goes up to Iran and nips off a corner)]
25.
by Lerc 22d ago
Berne convention countries plus a few more I guess, Maybe not Eritrea or Kosovo?
26.
by Lerc 22d ago
Well it's hard to name them because the names keep changing. There's a few that are just going for random letters, and simonw's stalker seems to pick a new name for the moment, but isn't really dedicated enough to match
27.
by Lerc 22d ago
You can do that legally as long as you do not confuse people that what you create might be an actual Disney product, or that you try to represent it as actually created by Disney. Both of those are protected by law. Learning from a Disney
28.
by Lerc 22d ago
>But from a social signalling perspective the price makes sense, it’s a way to show that you have lots of money. I think it's not purely an outward facing signalling. I remember attending a lecture by a former Apple designer on desi
29.
by Lerc 22d ago
On the topic of the sibling reply to this. There are quite a few new accounts that immediately go on attack on a few topics and individuals. I don't suppose you have any numbers of the ratio of IP addresses to attack accounts?
30.
by Lerc 22d ago
Can you offer any evidence that it is any less accurate than a page hand crafted by a human? I thought it provided an interesting tool for people to put things into perspective.

More ›