pull down to refresh
@optimism
724,299 sats stacked
stacking since: #879734longest cowboy streak: 144npub13wvyk...hhes6rk47y
123 sats \ 1 reply \ @optimism 12h \ parent \ on: You are not smart enough to make it. Neither am I. - David Shapiro AI
Right. Except now that you had attention, you blew it, because you literally had nothing to say - except begging (?!?) It took me serious effort to read after the fifth word.
I understand that you don't feel taken serious at times here on SN, but let's be honest, have you actually given thought to all this? You challenge this same thing over and over but you don't have any solution. How on earth is that possible?
PS: Do you really believe that Elon is a short-term operator?!? haha
Elon Musk kindly showed the Chinese how to build the best EVs
Suppose Elon reads this (he's known to lurk in places online so I personally wouldn't bet against it), what would you recommend him to do?
International law? Is that still a thing?
Only in countries that aren't consistently violating the agreements they once-upon-a-time made, back when the horror of WWII was still fresh. The main issue for those of us inhabiting parts of the world that aren't represented in the UNSC, is that those that are behave like a united front of bullies. Trump's idea of abolishing the UN isn't so awful if you realize that that would free up time by not having a lot of useless meetings that are better held without a boys club that anyway vetos everything.
For those among us that are represented in the security council: imagine how cool it would be if your human rights would actually be respected rather than pissed upon (and in some cases, right before you're being sent into the meat grinder to catch bullets for your oligarchic overlords.)
A lead in installed robotics! Ah! Why do you think that this particular race is so important? 1
Footnotes
-
Asking because, per the reports I linked you, which no doubt you read, you can see that Japanese and European top robotics manufacturers are selling to China. Especially for Fanuc this is interesting: China is their biggest market in 2025 (thus far). So although you're right that China has the largest installed base, even Chinese high-end fabs use foreign robotics, this is well known, and is only increasing. ↩
I don't direct the territory really, so I'm generally happy how it goes. The thing I was worried about most was triggering slop. But it looks like other territories have about the same amount of slop, because it's really just a select set of soulless milkers. It's not a concern right now, imho.
I'll keep on adding some emphasis on open models / research in my own posts and comments; to add balance to the big tech / closed source hype, because I believe that ultimately, humanity needs sovereign compute to stay safe from exploitation.
it would give China a huge advantage given their massive lead in robotics.
Idk mate... maybe @BlokchainB has more insights into this as our resident industry + AI stacker. I don't track industrial robotics myself, but why would the incumbent robotics providers still show growth if China has a massive lead? All I hear is that the current generation of Chinese robotics are extremely cheap, but I've heard at least some guys that ought to know (for example because they do process design for a living) say very recently that the reliability is still an issue versus Japanese robotics 1
But in technology development, planning design and manufacturing and productivity related to manufacturing AI combined with robotics does look to be a significant factor
Everyone is doing this, and have been for years. I think the interesting newer AI applications are on control plane and safety. See for example this nice product video from RA about their use of AI in continuous monitoring solutions, that can run off existing data as primitive as current drawing from engines.
Footnotes
-
In manufacturing, every second of downtime is lost money, even if your equipment is relatively cheap, as mfg contracts have actual hard timelines and you pack that to the max. ↩
I didn't remember before you mentionend it and I'm too lazy to do a search. Let's just re-label
repetition as cognitive honey, then it's a good thing, and stickier.Ah! Bitcoin is very competitive in payments, but you can't have a decentralized currency that you self-custody and have it as frictionless as something where you let someone else figure it out, unless that someone else gives you pain (like with your card, but it can be worse.) But like I say I'm dumb sometimes; to me, the value proposition of Bitcoin hasn't really been mass adoption (though that doesn't mean it has to be niche) but more a systemic solution that makes it hard for some a-hole to decide one morning to mess you up (despite what is being said about Luke today - the drama is not a feature.)
As for the comparison with AI, I'd argue that sovereign AI is more like Bitcoin, but hosted AI is more like Visa. It's just a matter of time for OpenAI to be the new Facebook and screw people over, and only those that have (or know of) no alternative will ultimately stick on there, just like every time I'm like "wait why are you on FB?" when someone talks about it. But then again, I'm dumb sometimes.
The slop is low. Increased a bit the past few weeks but nothing crazy.
So no reason to have high fee
I think it is far too soon to say what the effect generative AI will actually be.
This. Plus, I feel that the most interesting AI applications today aren't purely generative, but transformative. The most used application today is probably search result enhancement because every search engine has it, and second, practical digestion of information, per #1223147.
So while it's super-awesome that you can with 1 voice command let it reply to all your emails with nonsense, that is not the killer app. The killer app in software is almost always that what empowers humans, not what replaces them.
C&P error in one of my spreadsheets, for @hn, before
bot was an icon, lol. Fixed it.If Kafka is still using Zookeeper
Not from 0.11 and up I think - i.e. since 8 years or so?
Was NATS a bust?
Sort of. I found it very hard to use outside of a server environment, needs a lot of work to make it survive dumb things that happen on desktop environments like sleep. The out-of-the-box inconsistencies after wake were driving me nuts. And I wanted this to ultimately embed it, but it's just easier to use leveldb and custom build the at-least-once logic on top, so I've switched to that.
Kafka is because I plan to do a ton of server-side processing in the future so I'm now trying to get a working dev environment w/ it... on macos, lol. I should just connect to my dev server but I want to be able to work offline too so now I'm down the apple container rabbithole because I truly dislike docker desktop.
In collaboration with Stanford Social Media Lab, our research team at BetterUp Labs has identified one possible reason: Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers. On social media, which is increasingly clogged with low-quality AI-generated posts, this content is often referred to as “AI slop.” In the context of work, we refer to this phenomenon as “workslop.” We defineworkslopas AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task.
Very nice definition!
Each incidence of workslop carries real costs for companies. Employees reported spending an average of one hour and 56 minutes dealing with each instance of workslop. Based on participants’ estimates of time spent, as well as on their self-reported salary, we find that these workslop incidents carry an invisible tax of $186 per month. For an organization of 10,000 workers, given the estimated prevalence of workslop (41%), this yields over $9 million per year in lost productivity.
I think it's much worse than this if you're in a paper-crunching or management job, but then the 1h56m spent for dealing with each instance is quite high. Perhaps some are not noticed as slop and just acted upon?
FWIW I can attest to this as a former hiring manager in the US for an international mid-size company. What was (and I'd guess still is, looking at the stats) happening was that the large consultancy firms would get cheap labor from overseas under H1B because the commercial rate of a junior consultant (aka someone that doesn't know anything but has great education) in the US was/is 20x-50x that of someone in South Asia or Eastern Europe. So if you give them a 5x raise you'd make 4x more money from letting the person talk to the customer from the US instead of their home country.
We mostly needed to get experienced engineers (to make field service more robust, because it was hard to get people with proper experience in the US) over from Western/Central Europe, and we were often not getting H1B allocations - these would go to Accenture & co instead - when we needed them, so instead we'd get people to Canada where there was no upper limit for experienced engineers, and only managers (including me) were under L1 in the US. Then at least, if field support was in need in the US, at least they were relatively close, rather than having to subject them to 8-12 hour flight plans.