For anyone else wondering why the article ends in a non-sequitur: it looks like the author wrote about decompiling the Claude Code binaries and (presumably) discovering A/B testing paths in the code.
Presumably the idea is that you put the relevant parts of the list in your thesis. You need to convince your examiner that you understand the background to the original research you did, and a solid reference list (with supporting text in the introductory/background section of your thesis) is part of doing that.
Personally I did the references at the end and didn't feel like I suffered from that decision, but the key references in my particular area were a relatively small and well-known set.
Hmm, yeah. I mean you often see huge reference lists which always just makes me feel like the person can't possible be actually well acquainted with the stuff that's being referenced. So who are you really fooling? Seems all very performative, though I guess I understand the motivation
What are the factors influencing the US Navy's position here? Not enough small/cheap ships for this work? Too hard to defend against guerilla speedboat attacks?
No minetrawlers, the four US had were scheduled to be scrapped earlier this year. So if there's even a single mine you're playing russian roulette with hundreds of people on board
Probably heaps of various anti ship missiles that have been squirreled away with ranges reaching from few nautical miles to few hundred, just for this exact scenario, please keep in mind that you only need one missile to get through to cause dozens if not hundreds of fatalities.
Unmanned naval drones of various kind, not exactly ultra-high tech in this day and age.
And then there's the guerrilla speedboat attacks which means more missiles
Did I mention that one ship has possibly hundreds of people on board? The political system of the US probably cannot tolerate a military mass casualty event of that scale and spectacle. It's therefore just too risky to get anywhere nearby with a ship so all US navy can do is just lob missiles from as far away as possible, while hoping that this whole mess ends before US runs out of standoff weapons. And between Ukraine, Yemen and now this, the armament stocks aren't probably looking too good considering the meager production numbers.
This isn't a military decision but more a public opinion one. Should an American ship take a hit, have casualties, become disabled, etc it would put immense pressure on the administration to settle/end the war, even though on a military objective level it makes a lot of sense. This is a reality of the instant informational world we live in.
Like a lot of things, little about this war is purely bad or purely good.
If the Iranian regime were over thrown, that would be good for basically the whole world except the people actually operating the regime. So, if the war ends without that happening, then that's at least partly a bad thing mixed in with the good of, y'know, not having a war anymore.
Imagine the optics of a single destroyer/cruiser being on fire. It would shatter the myth of American naval power (some are arguing that this war already did that, which I tend to agree with).
Its like the issue with the Vietnam war. You need 100% perfect security, or its not worth it. If you are only 98% successful, you arent going to have oil tankers or any cargo ships even attempting it. A single failure every 2 months was a massive waste of resources.
While Iran still has fire control, these ships can be hit by shore-launched anti-shipping missiles, one way drones of even old fashioned shelling. Their "navy" was never even a factor.
Too risky, and doesn't make sense from a cost-benefit perspective. Iran uses cheap and disposable weapons that are also effective. If you think about how much a single US ship costs, and the political price of US service members dying, I think the picture becomes clear.
The decision of the US Navy to not provide escort services makes perfect sense and it is no surprise.
The only thing that is newsworthy about it is that this has exposed yet another lie of Trump, who at some point has promised that the traffic will not be affected, because USA will provide such escort services.
Protecting the economic interests of the US and allies by defending commercial shipping is well within the remit of the US Navy. This is a risk management decision with a side of optics, not one of scope.
> Shaming individuals doesn't seem to be productive or helpful.
I don't see how much support from history for that viewpoint. Some examples of positive societal change driven in part by shaming individuals: drink-driving, civil rights, sexual harassment, automobile safety, the slave trade, McCarthyism.
All those cases also have huge penalties or effective costs associated with them. Is there an accurate "shame first, then penalties came later" stand point?
Automobile safety in my life has only changed after fines.
Sexual harrassment still happens and doesn't seem to be helped by shaming someone as much as firing them. Though we often don't have the guts or legal backing to publically shame someone.
Eh, other people throw litter on the floor and rob elderly folks in their homes. Those people hardly ever get caught, but neither you nor I are are going to start copying their actions.
Amsterdam also reduced many roads from 50 to 30 kmph. Accidents have reduced by 11% and travel time has only increased 1-5%. That is less than one minute on a 20 minute trip.
I haven’t looked into this specific case, but most of the time the limiting factor is other traffic. You’re not traveling at full speed the whole time. If a lower speed adds 10 minutes to the average trip, but it reduces 9 minutes’ worth of traffic, you’ve only lost net one minute. A lower speed limit will often reduce traffic because the speed-up-slow-down behavior is reduced.
Personally, I have driven around the Netherlands a fair bit and this sort of thing does seem to be roughly true for the median case. It can definitely be annoying when the streets are empty, though. For those journeys you’re obviously losing a fair bit of time.
I have a long stretch of road near me that used to be 50 km/h and is now 30 km/h.
If you take that road during rush hour, yes - there's no meaningful difference in time spent. You'd be going traffic light to traffic light slightly faster.
The problem is that I personally use that road (in a car) more to pick up/drive my wife from and to early and late shifts at work than during rush hour, and this makes it take a significantly longer time. Like, I am not complaining at all, but it takes something like 20 minutes instead of 15, so like a 33% increase. And then again on the way back. But in the end being lucky with the traffic lights is still the main point.
Maybe people got so frustrated by having to drive at a snails pace that it became preferable and/or faster to just use other modes of transportation which cut down on traffic improving travel times?
It's how most of us are actually going to end up using AI agents for the foreseeable future, perhaps with increasing degrees of abstraction as we move to a teams-of-agents model.
The industry hasn't come up with a simple meme-format term to explain this workflow pattern yet, so people aren't excited about it. But don't worry, we'll surely have a bullshit term for it soon, and managers everywhere will be excited. In the meantime, we can just continue doing work with these new tools.
This is an opportunity to select some stupid words that you would like to hear repeated a million times. The process is like patiently nurturing a well-contained thing, so how about "egg coding"?
I havent quite dealt with "teams of agents" yet outside of Claude Code itself spawning subagents, but I have some ideas as to how to achieve it in a meaningful way without giving a developer 10 claude code licenses, I think the real approach that makes more sense to me is to still have humans in the loop, but have their respective agents sync together and divide work towards one goal, but being able to determine which tasks are left to be worked one and tested. I do think for the foreseeable future you will need human validation for AI.
I'm not sure there's going to be a term, because there's no difference from normal, good quality engineering. You iterate on design, validate results, prioritise execution. It's just that you hand over the writing code part. It's as boring as it gets.
In a couple of decades of work, I have never actually met anyone like Julius. Typically, I have found that those who excel at listening and presenting are also capable of understanding the technology at an appropriate level for their role -- it's not like this stuff is truly complicated, after all.
I have met quite a few people who are more focussed on the business than the technology, but those people tend to end up in jobs where the main problems aren't actually technical. Which, let's be honest, is the case in very many tech jobs.
oh man, I have met several Juliuses. one of them was my boss till he made an error as similar to the one the original Julius made, but unfortunately too late I had to leave the company earlier he made my life hell. now he is at another company, as long as he is at this company I won't apply there, if they hire him they have no place for me
I've seen it rationalized by saying you should be moving jobs every year or so, because if you're not doing that, then you're not growing. I've always thought of this as a sort of Julius coping mechanism. On some level, I think a Julius views a non-Julius as a stagnant old gray beard who rejects the "growth mindset".
To be clear: I've never seen people who follow this strategy contribute anything of value, and it's the biggest red flag on a resume. You learn and grow more by seeing things through.
I have met armies of julius at all levels. Id say 80% of people are julius and if u dont think so then i have some news for you.
It is always like this. Your ability to socialize will bring you further than any other skillset. The Kennedys for example manufactured their status by socializing. Industry is no different.
Hi julius! I kid and dont speak as if i am not a julius. Most people are in it for money for a house car family etc. they dont care about the job in so much as means to an end. That is julius but he took it further
that is not what makes a julius though. there are lots of good, competent workers who don't really give a damn about the job, and are just in it for the money, but they know their stuff and are genuinely working and delivering value for the money they are given while they are on the clock. what a julius dials up to eleven is the oft-heard dictum "fake it till you make it", only they are so good at faking it that it becomes their entire thing, not just a way to stay under the radar while they learn the job.
80% of the people are saying that this is highly complex software. We should not expect to serve more than 4 requests per second without a full kubernetes cluster backed by 27 pods, a cloud spanner database, and 200k lines of code.
> Id say 80% of people are julius and if u dont think so then i have some news for you.
> Industry is no different.
Based on these comments, maybe some self-reflection is in order, as it seems from the 80% comment that what you mean is that 80% of people are able to adequately communicate.
That number feels off by a lot to me. I think i can say i'm quite good at socializing, quite above average when comparing to people I meet and work with. I'd rate my engineering skills about average level and i have a firm dislike of fraud and of people acting to be better/smarter/faster than they really are. In my career I've come across managers of the julius type, as well of the narcissistic type, even a sociopath. I would estimate 10 to 20 percent of people are of the Julius type.
It was a subtle ref to the 80/20 rule in that most people likely oscillate between the julius and the useful. Some of that 80% are full time julius for sure.
I am not sure how one even gets 250TB/mo through a 1Gbps link. In any case, completely saturating your networking for the full month is outside most people's definition of "fair use".
Yeah but they still advertise with unlimited traffic.
"All root servers have a dedicated 1 GBit uplink by default and with it unlimited traffic"
https://docs.hetzner.com/robot/general/traffic/
HN user 'onion2k pointed out that doing this breaks Anthropic's T&Cs: https://news.ycombinator.com/item?id=47375787
reply