empiko

> The "AI voice" is everywhere now.

Maybe I'm going crazy but I can smell it in the OP as well.

I'm with you. I hope that the other posts will also pop up here.

> So it is just a case of older people pulling the ladder up behind themselves.

Is it though? I have a feeling that previous generations were simply happy with less. Now we are so connected and everybody wants what they consider the standard according to social media: huge house in the most prominent city in their country, N exotic vacations every year, meaningful job, etc. But this would be a pretty tall order even 20 or 40 years ago.

If this was the case, we would see R&D costs dropping for OpenAI. Not sure if that is the case.

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.

HackerNews - hnrss.org

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.

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.

There might be a lot of people interested in all kinds of small apps for their particular hobbies, interests, jobs, etc

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 papers with 50+ references for all kinds of claims and random semirelated work. The citation culture is really uninspiring.

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?

Yep, it is a thing in a handful of Western European countries.

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.

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 weakened, the bully comes and takes our lunch money. EU politicians should start taking realpolitik lessons, the rest of the world is playing with us.

> 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 feedback on theses texts and papers. I see people using AI to write their texts more and more, and I feel like my hours are now wasted on improving AI-generated texts instead of helping people hone their writing and thinking skills. Since I cannot constantly analyze and think about who actually wrote the texts, I am thinking about stopping my volunteering.

"Continue reading this post for free in the Substack app."

No thanks

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.

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 I want to do. It will then promptly copy the existing code into this new function and suggest to replace the occurrences in the code with it. Tab, tab, tab, editing is complete in a few clicks.

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 encyclopedia. Their costs are probably very strongly correlated with their number of visitors.

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 depends on you working hard and you have very little capital.

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.

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 UX team with N people, the team will make sure to generate workload for N people, probably even more.

On top of that, managing a huge redesign is a great career opportunity for everyone involved. The incentives are simply stacked in favor of doing redesigns for their own sake all the time. You need a clear minded top level manager to stop these kinds of ideas.

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 to filter the generated URLs aggressively.

IMO they just don't have any idea what data are actually copyrighted and are too lazy to invest in the problem.

This might very well be happening, but fyi, current LLMs tend to be biased in favor of traditionally disadvantaged groups.

At the start of the first meeting, you annouce that you need to leave at 1:50 and ask the meeting to respect that.

I wonder if they unionized in an attempt to save their jobs.

Well I don't see people running their web search locally, so I don't think they will run their own search+LLM.

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 years are probably hitting the practical limits of scaling for now.

I also feel like most insiders were fully aware of this fact, but it was a neat sales pitch.

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 very soon.