matusp

5 ms·

61.
by matusp 8mo ago
If this was the case, we would see R&D costs dropping for OpenAI. Not sure if that is the case.
62.
by matusp 8mo ago
I don't think the math is correct here. The 25% of positions that have been seen more than once represent more than 25% of the occurrences. Even if all of them would be seen only twice, you should already see them in 40% occurences.
63.
by matusp 8mo ago
HackerNews - hnrss.org
64.
by matusp 9mo ago
Just a side note, but for most of history, art was participatory. People participated in rituals, songs, dances, etc. It's only a recent idea that art is about consumption of what others have created.
65.
by matusp 9mo ago
I think it's a question of how polished the process is. Imagine a new fully functional app popping up in your phone after a single prompt. Then, imagine you can edit this app while you are using it by saying what you wish it should do.
66.
by matusp 9mo ago
Nobody in ML or AI is verifying all your references. Reviewers will point out if you miss a super related work, but that's it. This is especially true with the recent (last two decades?) inflation in citation counts. You regularly have
67.
by matusp 9mo ago
I do not disagree with the post, but I am surprised that a post that is basically explaining very basic dataset construction is so high up here. But I guess most people just read the headline?
68.
by matusp 9mo ago
Yep, it is a thing in a handful of Western European countries.
69.
by matusp 9mo ago
Alternatively, you can bet a few tens of thousand that your enemy is going to be alive in a few weeks and let the market handle the rest.
70.
by matusp 9mo ago
EU is in a tough place. Economically and security wise, we are in a bad situation due to the EU-Russian split, which was actually started by the US with their moronic NATO expansion policy during the GWB presidency. Now, when we are weakene
71.
by matusp 9mo ago
> As a maintainer many PRs now look like an insult to one’s time, but when one pushes back, the other person does not see what they did wrong. I have been volunteering as an advisor to various master's and PhD theses, giving feedbac
72.
by matusp 9mo ago
"Continue reading this post for free in the Substack app." No thanks
73.
by matusp 9mo ago
I expect that once one player adds ads, all the others will follow when they find out that the coast is clear and users are not leaving in droves.
74.
An Unfolding Scientific Revolution in Cosmology(economicsfromthetopdown.com)
28 points by matusp 9mo ago|8 comments
75.
by matusp 9mo ago
One thing that I do not see mentioned enough is how fast editing code is now. I want to create a new function to replace a functionality that appears in multiple places? I can just write the name of the function and the IDE understands what
76.
by matusp 9mo ago
Wikipedia is not even in the game you are describing here. Wikipedia does not need billions of users clicking on ads to convince investors in yet another seed. They are an encyclopedia, and if fewer people will visit, they will still be an
77.
by matusp 9mo ago
Farming is "easy" when you have your tech savings and the option to plunge back into a high earning career when shit goes sideways (with wise sage aura, cause you took care of some goats). Farming is harder when your entire family
78.
by matusp 9mo ago
Yup, 15 is just too many. I think that 10 is already pushing it, depending on how many projects are going on at the same time.
79.
by matusp 9mo ago
I believe that sentiment to some extent. As soon as you establish an org, they will keep generating projects for themselves. Almost no manager will tell you that their work is complete and it is time to downsize their team. If you have an U
80.
by matusp 9mo ago
I wonder how they handled it. Everybody's connecfing their AI to the Web, but it automatically means that any data AI has access to can be extracted by the attacker. The only safe way forward is to 1. disconnect the Web or 2. perhaps t
81.
by matusp 9mo ago
IMO they just don't have any idea what data are actually copyrighted and are too lazy to invest in the problem.
82.
by matusp 9mo ago
This might very well be happening, but fyi, current LLMs tend to be biased in favor of traditionally disadvantaged groups.
83.
by matusp 9mo ago
At the start of the first meeting, you annouce that you need to leave at 1:50 and ask the meeting to respect that.
84.
by matusp 9mo ago
I wonder if they unionized in an attempt to save their jobs.
85.
by matusp 9mo ago
Well I don't see people running their web search locally, so I don't think they will run their own search+LLM.
86.
by matusp 9mo ago
Scaling works, the problem is that it is practically impossible to scale much. There is only so much energy, text data, GPUs, etc. The folly of scaling is that we are living in a finite world. The huge investments in AI for the past few yea
87.
by matusp 9mo ago
The sota chatbots are getting more and more functionality that is not just LLM inference. They can search the web, process files, integrate with other apps. I think that's why most people will consider local LLMs to be insufficient ver
88.
by matusp 9mo ago
This is a natural response to software enshittification. You can hardly find an iOS app that is not plagued by ads, subscriptions, or hostile data collection. Now you can have your own small utilities that can work for you. This sort of per
89.
by matusp 9mo ago
It's not really a mystery why it happens. LLM APIs are non-deterministic from user's point of view because your request is going to get batched with other users' requests. The batch behavior is deterministic, but your batch i
90.
by matusp 9mo ago
Not convinced. There is an obvious value in having more food or more products for almost anybody on Earth. I am not sure this is the case for software. Most people's needs are completely fulfilled with the amount and quality of softwar

More ›