Ask your manager how they get humans to do those things, and copy that process.
So, get a bug tracker, track those bugs, tell Claude to pick tasks off it etc.
I'm not claiming this actually works, I've not tried it, I don't know how good it is for large brownfield projects, but that's the general sentiment I see.
I've tried it. It can work. My prompt was "Use the gh commandline tool to get the issues for the current repository, and work on them in order, with bugs taking priority."
Elsewhere there are steps for how to develop: 1. Create new branch for the feature you are working on; 2. implement the feature fully, thinking hard when you need to (toolcall think(low, med, high) switches the reasoning level);
Yeah, kinda, but the moment it crosses the "war" threshold the pain rapidly escalates to things like "I hope you weren't using the Texas refineries, because half of them are now on fire".
It's not like Europe isn't looking at the effect Ukraine has been having on Russia and going "hmmm, interesting". Obviously I'm still over-simplifying, anything I'm aware of will have been war-gamed do death and back.
May happen, but I suspect not in the way implied by that question.
Hardware is still improving, though not as fast as it used to; it's very plausible that even the current largest open weights models will run on affordable PCs and laptops in 5 years, and high-end smartphones in 7.
I don't know how big the SOTA close-weights models are, that may come later.
But: to the extent that a model that runs on your phone can do your job, your employer will ask "why are we paying you so much?" and now you can't afford the phone.
Even if the SOTA is always running ahead of local models, Claude Code could cost 1500 times as much and still have the average American business asking "So why did we hire a junior? You say the juniors learn when we train them, I don't care, let some other company do that and we only hire mid-tier and up now."
(Threshold is less than 1500 elsewhere, I just happened to have recently seen the average US pay for junior-grade software developers, $85k*, which is 350x cheaper, and my own observation that they're not only junior quality but also much faster to output than a junior).
* but also note while looking for a citation the search results made claims varying from $55k to $97.7k
Surprisingly, there do still appear to be people (or at least one person) here on Hacker News who think America would win and that the EU would simply accept being invaded and do nothing. Like the EU would not decide "fight or die, we
choose fight", nor replace the US with China in trade relationships.
This insight is… the interaction felt like reading a history book, or a novel with an obvious villain.
Feels unreal, like watching 9/11 unfold on TV felt a quarter century ago. Except I get to talk directly to the Other instead of watching translations of their words being quoted on bulletins.
I'd agree, the code "isn’t necessarily broadly coherent but is locally impressive".
However, I've seen some totally successful, even award-winning, human-written projects where I could say the same.
Ages back, I heard a woodworking analogy:
LLM code is like MDF. Really useful for cheap furniture, massively cheaper than solid wood, but it would be a mistake to use it as a structural element in a house.
Now, I've never made anything more complex than furniture, so I don't know how well that fit the previous models let alone the current ones… but I've absolutely seen success coming out of bigger balls of mud than the balls of mud I got from letting Claude loose for a bit without oversight.
Still, just because you can get success even with sloppy code, doesn't mean I think this is true everywhere. It's not like the award was for industrial equipment or anything, the closest I've come to life-critical code is helping to find and schedule video calls with GPs.
Interesting; as a British person myself, we don't get taught any of that about Norway or Denmark, not even knowing that they were once joint together in a union.
I'm not surprised. From a British POV it was a relatively minor part of a much larger conflict that Britain was done with when Napoleon defeated, and Denmark-Norway was for most practical purposes treated as "just" Denmark, since Denmark was the more powerful part of the union by far.
From the Danish and Norwegian side, Britain annihilated or captured most of the Danish-Norwegian fleet because Britain expected Denmark-Norway to enter the war on Napoleons side (as a consequence, Denmark-Norway of course entered, but severely weakened), and Norway was blockaded and faced famine from 1808-1814.
After the war ended, the Norwegian mainland was handed over to Sweden (Iceland and Greenland were also Norwegian at that point, but stayed with Denmark), but Norway took advantage of the process and passed a constitution and briefly went to war against Sweden to force a better settlement, resulting in a relatively loose union. So this whole affair had a very significant effect on the formation of the Norwegian state.
> The US economy is by far the world's largest and now dwarfs the Eurozone.
Nominal, Eurozone, yes.
But, being the reserve currency boosts the exchange rate all by itself. I'd argue that this acts as hysteresis, that it adds strength that keeps it a reserve currency longer than it would if there was no memory in the system. Therefore, if anything does induce a shock, the PPP rate is more relevant when considering who might displace it; this other currency (or currencies) would then also get the same hysteresis benefit.
The EU, PPP, is about the same as the US (30 T), and I'd argue that "the EU" is important measure for near-future stuff rather than the current Eurozone, because the EU has the no-specific-time-constraint preference to become all Eurozone… except for the bits that opted out. But also some more neighbours who opted in without being in the EU. It's weird.
China, PPP, it is bigger than the US, 40 T by PPP. Not quite as big as the gap between the US and India, but close enough I had to get the calculator out I can't eyeball the ratio on a linear graph: https://upload.wikimedia.org/wikipedia/en/timeline/jfgbd60rb...
> To answer your question: the US economy is going to last a very long time yet. So far it has lasted hundreds of years. Please provide a comparison to any other economy that has lasted so long and done so well. You'll be able to name two or three examples maximum.
You didn't do well for all of those hundreds of years, if you squint hard enough to ignore the great depression you get to about 150 years, which basically means about the same as every other industrial economy that didn't have a war in the middle split it apart. If you don't do that (because the great depression really sucked), the half of Europe whose national boundaries explosively reorganised, and also the Soviet Union, wave hello.
The USSR is an important reference, because basically nobody saw the collapse coming until a year or two before it happened. It was unthinkable.
> In the moment people tend to get hyper emotional, hyperbolic. They think something fundamental is changing. That's almost always nothing more than personal subjective projection of what they want to have happen, rather than an objective assessment of reality.
All true.
> Back in reality the US has survived and thrived through drastically worse than anything going on in the present. The Vietnam era was far worse both socially/culturally and economically. WW2 was drastically worse. The Civil War was drastically worse. The Great Depression was drastically worse. But oh yeah sure, the US superpower is about to end any day now.
How many of those occasions did the US refuse to rule out military force with its primary set of allies in order to seize land supposedly to keep it safe from a nation that's now 33% richer than it is? The Civil War was not a time when y'all were a big player on the world stage, it was when Europe was busy carving everything up into colonies.
At this point, I'd say terms rather than generations.
I mean, I'm old enough to remember people saying "Never Forget" about 9/11, but it's barely in any discourse at this point, and that was a single generation ago and had two major wars a bunch of PoW scandals, war crime scandals that led to Manning, and domestic surveillance that led to Snowden. And yet, despite all that, I've only heard 9/11 mentioned exactly once since visiting NYC in 2017, and that was Steve Bannon and Giuliani refusing to believe that Mamdani was legitimate.
So, yeah, if Trump fades away this could be forgotten in 8 years or so; if this escalates to a war (I'm not confident, but if I had to guess I'd say 10% or so?), then I see it rising to the level of generations.
It's different. 9/11 was an outside foe, which was dismantled by US forces, and its leader was executed. America "won" against the perpetrators of 9/11 in the conventional sense.
You cannot defeat MAGA the same way: the "enemies" are among us, and they aren't going anywhere.
From my point of view as a European asking if myself if or when I will be able to trust the USA in the future, the Taliban is to Afghanistan as MAGA is to the USA.
You're the outsider, to me. The pre-9/11 Taliban were seen as "kinda weird but we can do deals, oh dear aren't they awful, never mind", the post-9/11 were not even worthy of talking to. The USA is currently in a similar "pre" state, an invasion would make it a "post" state.
I'm saying "never forget" fades. That's a human condition we all share.
I mean, I live in Germany these days, and this country absolutely got the multi-generational thing, and I'm from the UK whose empire ditto, but… the UK doesn't spend much time thinking about the Falklands War and even less about the Cod Wars.
Nobody disagreed with that it eventually fades, they were all saying it is going to take decades. The consequence of 9/11 of was mostly TSAs, following the USA into wars and the erosion of privacy at the mention of terrorism. The first and the last are still ongoing and I think the current US admin is still using the latter as a narrative, the second one may come at an end currently, because the USA is trying to use it against its (former) allies.
What you describe is called "to historicize an event". The WW1 has been historicized by WW2 (some argue it's the same war). But not even WW2 has been historicized yet (at least in Europe) and it already ended 80 years in the past, so I doubt an atlantic conflict is going to be forgotten in the next few decades.
Edit: I originally linked to https://en.wikipedia.org/wiki/Historicization, but this does not describes what I mean. It is weird, because the supposed German equivalent does. The German article is about a concept from the science of history, while the English article is about a literature concept.
Aye, and thanks for the link, will read the german version as per your edit.
> so I doubt an atlantic conflict is going to be forgotten in the next few decades.
If it gets to one, yes. Was writing late at night, so sloppily, sorry about that.
Right now, I think we're not that far gone yet. Absolutely agree it becomes as you say if it becomes hot war. Not sure about which step between will be the drop that overflows the bucket.
If we don't reduce conflict to mean military conflict, then I think there is definitely some diplomatic issue ongoing.
> Not sure about which step between will be the drop that overflows the bucket.
True, this is kind of the open question, because the EU both needs to be the adult in the room and deescalate, but also can't do compromises with territorial integrity otherwise it has already lost. This will of course have an impact on the "time to forget".
But I don't think if there is a uprising today in the US, Trump and the whole admin is gone next week and they improve their constitution, that the whole issue will just be forgotten. The whole pro-, neutral- or even contra USA debate has been ongoing for decades know. For example the trade deals aren't exactly concordant with EU law (https://en.wikipedia.org/wiki/Max_Schrems#Schrems_I) and the USA has been boycotting multilateral institutions, that the EU wants to have authority. I mean it is new that they openly sabotage the ICJ, but that they have the capability to do that is not.
Yeah, one thing the EU could do that wouldn't hurt them/us (much) would be to stop bringing up fake replacements for the data sharing agreements that get shot down.
The damage would mostly hit the top performers of the US stock market (amongst others) while not damaging the EU as much.
It'll probably be tariffs first though, followed by the ACI if things get really bad.
There's how the people in general remember, and then there's how the politicians and the institutions remember. If nothing else, the changes in institutions will have effects reverberating for decades, with the most obvious institution being the military in each country that expected to fight a war under a NATO umbrella with an American general in charge.
If I'm a German or French or Swedish officer, especially if I'm suddenly in Greenland, I'm going to be thinking hard about the changes to come in the next few years so that they're not all dependent upon a friendly America. If nothing else, they're all getting ready now to operate without any Americans in the loop, since it might be Americans they're fighting. That means the entire NATO command structure, which presumes American dominance of it, is now an obstacle to avoid rather than a resource to share. Every PM is asking the head of their air force if they can fly their F-35s without the Americans knowing about it and possibly shutting them down remotely.
There's a story going around today in French newspapers about how French and Ukrainian intelligence fed US intelligence some false strategic info to see if it ended up in Russian hands, which it did within days. Now Ukraine is consciously breaking its relationship with US intelligence because it can't be trusted, while getting closer to French and German intelligence. I suspect that the UK is also carefully looking at what's shared via the Five Eyes and decided what it can/needs to withhold.
So, get a bug tracker, track those bugs, tell Claude to pick tasks off it etc.
I'm not claiming this actually works, I've not tried it, I don't know how good it is for large brownfield projects, but that's the general sentiment I see.
reply