pull down to refresh
@optimism
729,038 sats stacked
stacking since: #879734longest cowboy streak: 145npub13wvyk...hhes6rk47y
11 sats \ 0 replies \ @optimism 54m \ on: Stacker Saloon
Woke up early. Got triggered by #1237375. Now offline processing ILO's workshop on Artificial intelligence and jobs.
2025-09-25:
- Day 1, morning
- Day 1, afternoon
2025-09-26:
- Day 2, morning
- Day 2, afternoon (scheduled as of writing)
We're often worried about
surveillance law, but perhaps we should be more worried about surveillance > law.The second golden rule of batch migrations is: archive all the intermediate data (tables) somewhere. This helps you troubleshoot in the future, in case something went wrong.
This is a remarkable piece that in my personal opinion, hits a bunch of nails right on their heads.
From years of working alongside governments, multilateral bodies, and civil society, I have seen how often ambition is lost in the machinery of politics. That is why this moment, fragile as it is, merits special attention—and maybe even a bit of hope. In this case, nations recognized that no single country could govern artificial intelligence alone, and that recognition created the space to begin building lasting institutions for AI governance.
The reality we must confront is that for years, our debates about AI have been dominated by hype and fear, recycled narratives that misdirect our imagination and our policies. The UN’s resolution represents the first attempt to break that cycle by creating institutions that can anchor AI in science, evidence, and cooperation. If they succeed, they can create a new narrative of AI: one that serves public purpose rather than amplifying unjust profit or panic.
Too often, we tell the same scary stories: an evil mogul in his tower building AI systems no one else can control, a machine that outgrows its makers, a gleaming future where technology erases our flaws. Each carries a fragment of truth, but together they obscure the realities already shaping human lives. Narratives like these shape policy and investment, while the most consequential applications are too often ignored.
The author is the President of the McGovern Foundation 1 - which I didn't know existed, until today. Love to see that there are more tech-optimists, bigtech-pessimists in this world.
Footnotes
-
McGovern was the founder of IDG (Computer/PC/Mac World, "for Dummies") ↩
gather on short notice at a Marine Corps base in Virginia next week
Imagine being so retarded to concentrate your leadership in a single location and actually not keeping it secret.
I'll write something up when I'm done testing and tuning (still haven't started, sorry) for self-validation/broadcast. Re: LN node. I've never approached it like that, running everything on one Pi, but can have a look once I'm done w/ the bitcoind.
Geopolitics / global conflict. Not of the economic kind, of the my fellow human beings are getting killed because of some dumbass overlord kind.
The remedy? Work hard and after work smoke 1/8th of a blunt and pass the fuck out.
PS: which aspect of AI? The Skynet aspect or the Sam Altman aspect? Both are stuff of nightmares.
I appreciated flaco's take on this from yesterday
Same - I had a reply but I didn't like how it sounded so I'm still drafting, lol
the entire relay network can be the same .. zero conf won't be reliable.
The network cannot be the same because there is latency between nodes, thanks to the awesome principle of the speed of light, processing requirements, memory latency, utxo db lookup speed... and so on. Therefore zeroconf cannot be 100% reliable, because even if you see everything, you cannot know what is included in a miner's template that got distributed to all these ASICs, or maybe you only see it after a block was already found... So even in a tx replacement fight, you need luck. It's a race.
I'll spend a while on it and see where I end up
Check out Replacement Cycling - that's a nice thread to unravel 1
Footnotes
-
Note that I don't really want to be discussing stuff like this in all its glorious larp-cases in the open if you don't mind - the defense is always aggressive replacement / rebroadcasting and mempool monitoring. And being aware of what you're using your node for. ↩
I should have said "relatively new" - 50 years minimum horizon, lol.
What do you think its primary purpose should be?
Disclaimer: never listen to a retard like me!
But since you are asking, I'd say that it depends on use-case.
- No mempool. I agree with Gmax' argument that
blocksonlyis a perfect mode, for "leaf" nodes 1. Also, I feel that while a properly running relay node may help, a badly configured one or low-capacity node doesn't 2. If you're just looking to do your own validation of (low volume) utxos, no mempool is the best mempool, as you anyway don't want to zeroconf. - Knowing what's going on with your LN channels (or other pre-confirmation risk monitoring). Like I said above, for watchtowers you want to know as much as you can, even stuff that is unlikely to get mined 3. This is technically more permissive than being a reflection of what gets mined, especially when we see pool / template decentralization.
- But we do need relays also without LN, because hub-and-spoke would be bad for centralization. For this, I'd say the mempool is basically your pre-validated set of transactions that you pass around, like a hot database of what you want to relay to others. This is the use case where - although not necessarily useful from where I'm sitting - filters could make sense if an operator feels strongly about it. Since someone filtering important transactions is a threat scenario that needs to be minimized no matter what, it shouldn't be too much of a problem when other operators do this.
- Edit because I forgot: And as a stage for mining of course. I wouldn't filter my mempool if I were a miner/pool, but I'd definitely look to filter my inclusion rules.
Footnotes
-
it would be useful to have (degraded?) smartfee work when a node has only the block inclusion statistics, i.e. estimate based on what was mined only, without tracking confirmation time of seen->mined. I'm still building some test scenarios based on a discussion we had about this on SN a week or so ago and have promised to share anything I think can be improved (but I'm slow, sry.) ↩
-
In the past Ian Coleman had a simulator (pre-segwit, pre-compactblocks and now discontinued?) that illustrated the impact of more nodes on the network in terms of hops a tx would need to travel and what kind of data use we'd look at network-wide, which was interesting to play with. ↩
-
I've had the temptation to make my watchtower
allownonstandardbut since there were little/no peers with that and I felt there may be some risks I would miss and then not catch (don't want to be banned from all peers), I instead ran with a maximally permissive policy through config and a much-larger-than-default mempool (had thought of using librerelay instead.) Either way, this is why I feel configurability is very important. Someone may need it for a reason no one thought of. ↩
I think the problem is a little more profound than ads alone though. OpenAI is showing to be a bit of a predatory organization, like Google, Facebook and TikTok before them (but we only worry about it when it's Chinese, lol). Interacting with their platform is a vote to be abused in any way they legally can, to improve their bottom line at diminishing benefit to the customer over time. For most people this isn't a problem until it's too late, but it's always good to recognize potential issues and ask yourself "what am I doing?"
This critical thinking is what the hype+fomo has successfully killed, and when the masses will fully turn against chatbots, and AI in general, because of the corporate players that have exploited them, Sam will be ultra rich and find new ways to exploit the masses, just like Zuck does. However, this happening will probably be good for the tech.
Okay. The "contribution" thing really depends on how well your node runs, how good your uplink is, how much people sync off you, and so on. In general, small, slow nodes are a burden to the network, so make sure to keep learning how to help the network best.
For actually verifying your own blocks and posting your own transactions, it's of course a great use-case.
Core attempts to model what will get mined.
I still feel that this is a bit new-ish of a narrative though, afaik it was b10c that started analyzing
cmpctblock reconstruction at some point. It is okay that the narrated purpose of the mempool shifts but perhaps this is the difference between what Core does and what Knots aims for, and all that is needed is coexistence.I'd argue that:
Coreis heavily LN-influenced and LN benefits from as-uniform-as-possible mempools between the sum of all miners, and your (watchtower) node. However, since there is no consensus over mempool, any limiting policy doesn't work (for a watchtower). LN is basically a zeroconf protocol, see this 2015 email on how policy messes with zeroconf, and then think about how Peter could analyze what happened with his modified nodes (2 replies below in that thread). The idea then becomes: Not relaying any transaction will be bad for other watchtowers, so as a community, the greatest value is the dissemination of transactions, so that everyone's watchtower knows everything.Knotsisn't LN-oriented at all, but instead focuses on L1 and community values. Nodes relay transactions and the operator is in charge of what to relay, and there are people that do bad things, such as shittoken mints, that make the network more expensive to run (and make it more expensive to transact.) The idea: It's in everyone's interest to try to keep this shit out, no matter the cost.
These are the opposing views. No ridicule is needed. Both views are valid and funnily, if we give up on the uniform mempool idea, and optimize implementations for the usecase we desire, yet can bring up the intellectual honesty that not everyone is in the same camp, then there doesn't have to be a war, at all, if people get their dumb doomscrolling butts off X and instead spend time building tools for their community to co-exist with the other.
The polarization is the problem, not the different ideologies.
This means that the chainstate could not be fully verified and is not specific to Knots. I've had this happen with Core on machines with:
- Bad RAM
- Bad disk
and it's really hard to reproduce for others. See also #1060416
If it happens again:
- Run memtest
- Inspect SMART reports for your disk
Let us know what it says if it comes this far - may make sense to do this regardless of whether the issue persists.