Transportation influences urban development. That is why most houses have a garage. There is no such thing as private transport (streets are public). Transportation has been heavily centralized since the New Deal. The bicycle was okay for most people living in cities in the 30s, now it is not because the government has favored the car infrastructure over the last decades. I think we need to start with not letting government develop their big infrastructure projects which are not resilient. Advocating for the use of bicycles might make sense in some places yet bicycle infrastructure is required.
Where I live there is plenty of bike infrastructure. I and many others don't use bike for transportation because of crime. Homeless steal bikes and parts of bikes if they cannot defeat the lock somehow. Recently a cyclists got killed in a "bike-jacking". People even get bikes stolen from their balconies on the 2nd floor. Reign in crime if you want people to use bikes more.
This is true. But it does not negate the comment you are replying to. Once you introduce kids into the mix (esp infants) - this whole narrative falls apart quite quick, ditto for elders/people with disabilities. Bikes, public transport are not a substitute for the vehicle.
I do agree that the vehicle should not be the default transportation even if I do consider myself a "car guy".
Also, federal highways are partially a national security issue, and are designed for quickly moving military equipment across otherwise isolated areas. Guidelines for federal interstates are specified jointly with the DoD to ensure that military transport can fit under bridges, and that bridges can support their weight. Industry is the other most important user, while individual consumers/families are the least considered users.
Everyone always assumes that individual choices and consumer behavior drives this stuff, and then they wonder why nothing changes even though we all started using reusable tote bags and LED bulbs. Stop blaming the consumer!
(The DoD is the largest institutional polluter in the world, by the way.)
That is very interesting. It is funny to see how influential the federal government has been on society, infrastructure and other areas of life. Specially considering that some people opposed to it during the confederation period because they saw it as another centralized authority (anti-federalist papers).
Trains are cheaper per mile but are less flexible and easier to sabotage. They are also important but there’s a reason that every country with a powerful military maintains both options.
I find it very useful for code comprehension. For writing code it still struggles (at least codex) and sometimes I feel I could have written the code myself faster rather than correct it every time it does something wrong.
Jeremy Howard argues that we should use LLMs to help us learn, once you let it reason for you then things go bad and you start getting cognitive debt. I agree with this.
I think LLMs are helpful for understanding code. I used to spend like an hour trying to find where something very specific was made, and now I can just ask an LLM and it finds it right a way and is able to explain how the code works. This is probably the thing that has saved me most of the time.
I've never learned so much and so fast as I do with LLMs.
1. Most of the learning before, especially technical related involved a lot of google searching for the information I needed. LLM here removes a lot of the friction and boring parts of the process.
2. At work I can leverage LLMs for some very mundane tasks, again, mostly related to information gathering. There was a time when I needed days to connect the dots in some very convoluted code written by your average developer. Even more to figure out the purpose of choices and how they connected to the business domain, often in situations where the relevant stakeholders and people with the know-how left the company. This kind of work, made of tons and tons of paper pages with my notes would sincerely exhaust me. And this kind of work has been the bulk of my career because coding was never the hard part. On this LLMs are increasingly better. This leaves me a lot of energy more to actually investigate the overall architectural decisions and technical details both of my projects and their dependencies (which have never been as easy to traverse).
3. Since I am less mentally exhausted (the only way to get mentally exhausted with LLMs is if you're "half vibecoding" so producing tons and tons of code which you are actually thoroughly reviewing) I have way more space to dedicate to learning. I do it both by practicing manual coding stuff for fun or editing the things I don't like in the work codebases I see, or by doing more katas on codewars or leetcode exercises. Also, I end up just in general asking more questions I would've not made just out of sheer curiosity and often learn a lot of things that suddenly "click". Another thing I do is way more spaced repetition exercises on topics I care (such as the many odd things you can learn with a language like C or metaprogramming coolness you encounter in Ruby and similar) on Remnote.
Honestly I don't get how you can learn less by having such a tool that removes so much friction.
But of course, if every AI naysayer conflates every LLM usage with vibecoding and with delegating the thinking and reasoning to LLM messages then sure, they are a disaster used like that. But that's on the user, not the tool.
The problem with LLMs is it doesn't bring structured learning or a curriculum. So it's great for filling in the gaps, but books and formal education are better at getting you a solid foundation in a field.
I agree with the point that learning requires work. In general, everything worth doing requires work. This is one of the things I often have to remind myself, otherwise I spend the whole day 'learning' and I just read a bunch of stuff online that I then forget, instead of trying something out which I actually will learn and understand.
I agree, though, you can still work toward understanding using an LLM (and take it from a skeptical) by, e.g., using them as challengers to your ideas.
That said, I think it requires a lot of self discipline and should be complemented with other methods and sources of information to be useful. As a teacher, I really try to prevent my undergraduate students from taking the easy road of using LLMs to solve every easyish problem I give them to *learn*. Sure, they did the homework but most of them did not learn anything while doing it and they finish their first year without having learnt anything actionable regarding computer science (observe that I use a different approach with students from other areas, though I still think it is good to spend a few days without relying on LLMs).
I often use a sport analogy to land my point which works with them, so let me share it here. If you want to learn how to run a marathon and drive 42km every day, then you are certainly (hopefully ?) a better driver but nowhere near to running a marathon (fortunately, no one has yet challenged me with the fact that running a marathon is way less useful to get a job than driving).
The robotics thing to replace caregivers misses the point that elder people also want connection. Yeah, it might free caregivers but still we will have a loneliness epidemic. I think this is more related to the desire for progress which is the backbone of modern life (you see it politics, school, your family, etcetera). This, I believe, has been slowly replacing the social glue of societies like religion, public space, play, chatting, etcetera.
As someone else mentioned here: not all technical debt is created equal. I agree, sometimes the problem are changing requirements, etc. But it is also true that there is technical debt caused by developers who don't take the time to properly design features and will simply implement the first thing that came to their minds. I agree with the author, this kind of technical debt is caused by a mediocre attitude which often propagates to all the team if there is no one that calls it out.
The more interesting discussion to me is: how do you solve this problem once it exists in a team? I guess there are many approaches, but I tend to think that 'lead by the example' is the best you can do as an engineer, but a top-down approach might work better which is what happened at Microsoft when Satya Nadella became CEO.
perhaps that was involved, but one thing clearly purposeful was people were seriously filtered for particular skills and personality (apple fit it was called back then), which created groups where individuals had unique skills and collectively the group members would naturally want to collaborate. it worked great.
(as an aside, this contrasts diametrically with Amazon, where i worked for a year for healthcare, not needing to because of Apple years' savings, but after a genomics startup i had joined ran out of funding, and wanting a new challenge; there skilled engineering types were presumed to be fungible assets for (not kidding) at least 7 layers of do-nothing bureaucrats making huge salaries...they could survive because sales on the amazon store extract something like the 30% royalty to amazon)
> "I expect us to go back to extending our agents with the most accessible programming language: natural language."
I don't agree with this. Natural language is so ambiguous. At least for software development the hard work is still coming up with clearly defined solutions.
There is a reason for why math has its own domain specific language.
Natural language can be bent into arbitrary precision. Write something, then enter a read-rewrite-reread loop as the devil's advocate (this is key) until it stops being ambiguous or having multiple conceivable interpretations.
Yes with English this process can be a pain in the butt, until you get the hang of it.
The problem is that it's very hard to anticipate all possible edge cases. Programming languages force you to do a lot of that work up front, English doesn't. It's the difference between writing Javascript and writing Typescript, except orders of magnitude worse.
The problem is, what's ambiguous or precise is subjective. Your devil's advocate needs to reflect all of the possible readers, and that isn't possible.
There's a good reason we use jargon in professions, or more constrained and less ambiguous languages for maths/coding
Was a pain to set up, but you can score the context completion and then if the score is under 98% or something, “ask” clarifying questions of the requesting agent or person or system
You’re never going to make a nontrivial statement in English that you couldn’t find two people who wouldn’t perfectly agree on its meaning. Or probably even a trivial one. Sure, at some point you can say “no, you’re clearly misinterpreting what I’ve said” or “you’re inferring something that wasn’t implied”, but English doesn’t have a formal spec or a reference implementation, so that’s kind of meaningless.
I was thinking the same thing. Maybe is that at the end the author seems to imply that agentic AI will work simply because models have become better regardless of the way we make them agentic (i.e. MCPs, skills, etc).
