abeppu

3 ms·

1.
by abeppu 5d ago
I have not looked carefully but it seems like this is over-promising on avoiding catastrophic forgetting. The "trunk learning rate" is set at 0.1x the learning rate for the experts, so learning on different subjects disproportiona
2.
by abeppu 6d ago
Even if there were as many positrons as electrons in the universe, I don't get the time-reversed character at a macro level. When leaving antimatter out of the picture, I dimly understand that the arrow of the arises from entropy. If t
3.
by abeppu 7d ago
I like that one of the successors to .yu was .me Combined with .it, not far away, one relatively small region which doesn't speak English has had all the best punny English pronoun TLDs.
4.
by abeppu 10d ago
Eliza wrote coherent English sentences.
5.
by abeppu 11d ago
I think this is a great direction -- for some kinds of users. And this makes me wonder if the 'vs' framing is misleading. Yes, I think it's a mistake that many organizations are cramming LLMs inside of automated pipelines whe
6.
by abeppu 12d ago
But isn't not-for-profit news media also struggling?
7.
by abeppu 12d ago
Or the quality of tokens will improve such that I don't have to ask model A to critique and summarize tokens from model B.
8.
by abeppu 14d ago
I think the slow down / pause vs acceleration framing is broken b/c it assumes there's basically only one forward direction which is to big and general models. Why should that be the case? What if instead of building one big
9.
by abeppu 15d ago
> I understand that technically, it's a project that can be used for defence No, it's a project commissioned by and built for defense needs. It's not that someone at the DoD saw how to creatively re-use an existing civilia
10.
by abeppu 15d ago
You should read the actual post? For Microsoft, the first and largest contract mentioned in the table just below the fold on that page is for an IVAS system. Microsoft had been basing this off their HoloLens system but was a distinct produc
11.
by abeppu 15d ago
Thanks for pointing that out; I was not aware of it. At the same time, when I helped with research in a cognitive psychology lab in school (much closer to the present than to that law), our PI said our funding came partly from a grant from
12.
by abeppu 16d ago
Ok you say rockets and space were "worthwhile" and missiles are "a different question". But of course the space race proceeded from work on missiles. ICMBs are rockets that go high and fast and then happen to complete le
13.
by abeppu 16d ago
I do think that for those of us who got into tech in the last couple of decades, the history can be rather obscured, but from the Fairchild semiconductor days, the military was a main customer, right? I think there's a real question: w
14.
by abeppu 16d ago
> Self-driving cars will never be safer than human drivers. You're responding to something that actually brought data to the table to answer the question, though one can quibble with details. Yet you're saying that self-driving
15.
by abeppu 16d ago
I do think that currently we're at the stage where Waymo is trying to convince most of the world of the _quality_ of its offering, so for the moment, things are well aligned. But in 5 years if there are multiple self-driving offerings
16.
by abeppu 26d ago
In my neighborhood there is an optometry place with a dark blue lighted sign. At night, it looks fuzzy, and I have to assume they knew this would be the case. People walking by squint at it and inevitably think "maybe I _should_ get my
17.
Tech giants urge global response to AI cybersecurity threats(dw.com)
2 points by abeppu 29d ago|0 comments
18.
by abeppu 29d ago
I think we gotta get people to see past the false dichotomy: part of their increased cost of living is due to climate change: - your home insurance is more expensive b/c fire/hurricane/etc risk is worse in your state - an
19.
by abeppu 1mo ago
> I strongly believe that video games have a higher ceiling for storytelling quality. There are several reasons underlying my newfound evangelism: a larger built-in audience, a more robust market with hundreds of games appealing to many
20.
by abeppu 1mo ago
I think also it's extremely hard to fairly judge high level decisions whose outcomes may not materialize for a very long time. This makes it genuinely hard for humans to get better (and one gets the impression that most of them don&#x2
21.
by abeppu 1mo ago
Wait, wait, gerrymandering messes with the House, but how does gerrymandering impact control of the Senate? Don't senators need to be elected by their whole state?
22.
by abeppu 1mo ago
> If all this sounds like wishful thinking, consider the polar case where housing deregulation drives the price of housing to zero. Yes, this wipes out the value of whatever housing you already own. But it also means that you can move to
23.
by abeppu 1mo ago
> Worse, it, to my ear, very clearly is written in a style that's extremely reminiscent of LLMs. In that way, it directly invokes that feeling of inauthenticity when he claims authorship. I get that and also find the style annoying.
24.
by abeppu 1mo ago
So, maybe this is not quite a fair analogy but a lot of famous modern artists don't do a lot of the actual execution of their work. If Jeff Koons relies on a team of artisans and technicians to make a balloon dog, and if Dale Chihuly d
25.
by abeppu 1mo ago
Why should intelligence _other_ than motor-planning be tied to the degrees of freedom of their movement? - snakes have a high number of degrees of freedom but aren't especially bright - squid have 2 more limbs than octopuses and aren&#
26.
by abeppu 1mo ago
So their photoscan bodyfat % has a lower RMSE in predicting (DXA) body fat than their BIA bodyfat % ... but what BIA estimation did they use? I've been told that not all BIA is equal. They tell us: > Fitbit Sense 2 was used to measu
27.
by abeppu 1mo ago
> It has been exceptionally well preserved because it was reused as a Roman-era paving stone on a colonnaded street The sculpture is very 3-dimensional. I don't get how it made a good paving stone, or how being used as a paving sto
28.
by abeppu 1mo ago
So the 1970s oil crisis caused a shift that included fuel efficiency rules. For like, the whole time I've been a software engineer, memory footprints of a lot of software have crept up, even when they aren't really doing more. The
29.
by abeppu 1mo ago
You can quibble with the specific projected savings, or the projected lives saved. But it kinda doesn't matter? If we can have net savings of half what this paper said, or a quarter, and still have more lives saved, that would still b
30.
by abeppu 1mo ago
Why do you expect an AI developed in the same economic system would optimize for your preferences rather than their revenue? I think this only works if people can run, fine-tune and update their own models -- otherwise the giant AI vendors

More ›