>Human-driven research is also brute-force but with a more efficient search strategy
No it's not. Is there anything to back that up? There's a creative aspect to human research that I've yet to see with gen AI. All it does is regurgitate stuff and get some "new" ideas via the latent space of the distribution it models. But a generative model cannot by definition create anything new. Just estimate its data well enough that it can sample it well enough to fake novelty.
OBBB removed any fines for violating CAFE standards. They still exist technically, but it'd be like getting a speeding ticket but the fine is always $0...
CAFE killed small trucks in part, tariffs in another part, but US manufacturers are the real reason small trucks are dead.
US manufacturers want margins, and they're not getting margins on little, efficient cars. They get enormous margins on gigantic trucks that start at $55,000. Have you noticed that all the sub $20k cars went away from all the manufacturers around COVID?
Ford makes the Maverick, which is a small truck. They were priced very reasonably at release, at $19,000 or so. However, Ford didn't make very many of them, and the ones they did make got up to $15,000 over MSRP from the dealers, who scalped them. Why would Ford want to cannibalize their pricy gigantic trucks when they know that they can get their $50k asking price because there's nowhere else for people to go?
>Why would Ford want to cannibalize their pricy gigantic trucks when they know that they can get their $50k asking price because there's nowhere else for people to go?
Why isn't Ford worried that Chevrolet, Toyota, Ram, or Nissan will bring back a small and cheap U.S. built pickup? Is that because all manufacturers are afraid of cannibalizing their more expensive offerings? Are they all colluding? Or do not many people want small pickups? I guess if the Slate becomes a breakout hit, we'll know that people really want the smaller pickups.
Neither GM, Chrysler, or Ford wants to hurt their expensive offerings. Toyota and Nissan have less expensive offerings, but can't bring them here because the tariffs make them much less margin, and the CAFE standards kill the rest off.
The Chicken tax didn't kill the domestically manufactured Ranger and turn the Colorado into the huge thing it is today.
CAFE killed them too. You can't have a small vehicle that gets fuck all MPG because it's built like a tank to do work. You gotta have a bigger one that gets slightly worse MPG but has a way huger footprint in order to make the math math.
This didn't just kill compact pickups for 20yr. It also killed the Chevy Astro (the most "fullsize work van" of the minivans) and why you'll never see a car with a giant overhanging cargo area again.
That’s not really sufficient explanation due to vehicles manufactured in the USA, CA or MX being exempt, and yet there are no small vehicles being made and sold in the USA in any large volume (despite clear demand).
My understanding is that this is due to fuel regulations being enacted by size and weight where it’s simply easier to make bigger vehicles.
I would also love to go back to Geocities style web interaction, but the medium is the message, and the way the Internet has evolved as a medium means that people don't naturally interact with it in a way that supports regression to that era. Attempts to force it like neocities have a hyperreal quality to them.
>Trump Says He Will Have the ‘Honor’ of ‘Taking Cuba’
>President Trump’s words came amid a nationwide blackout and as a top Cuban official said his country would move to open the economy to foreign investors.
It's strange - if we're so confident that our ideology is superior then wouldn't we welcome a small neighbor on our doorstep trying something different as a sort of experiment expected to reaffirm our views? Celebrate them the way you might a silly child, occasionally support them, all while pointing to them as an example of why we don't do what they're doing.
That was an awful long time ago and doesn't reflect the current situation.
> Cube prevents its citizens from leaving.
What's your point? If they suddenly changed that policy do you believe we would immediately walk back the sanctions and the oil blockade? That isn't how it looks to me.
The only way I can think to interpret current US policy is either one of petty insecurity or else an attempt at coercive commercial exploitation.
Haven't seen this mentioned yet, but the worst part for me is that a lot of management LOVES to use Claude to generate 50 page design documents, PRDs, etc., and send them to us to "please review as soon as you can". Nobody reads it, not even the people making it. I'm watching some employees just generate endless slide decks of nonsense and then waffle when asked any specific questions. If any of that is read, it is by other peoples' Claude.
It has also enabled a few people to write code or plan out implementation details who haven't done so in a long (sometimes decade or more) time, and so I'm getting some bizarre suggestions.
Otherwise, it really does depend on what kind of code. I hand write prod code, and the only thing that AI can do is review it and point out bugs to me. But for other things, like a throwaway script to generate a bunch of data for load testing? Sure, why not.
I've been tasked with code reviews of Claude chat bot written code (not Claude code that has RAG and can browse the file system). It always lacks any understanding of our problem area, 75% of the time it only works for a specific scenario (the prompted case), and almost 100% of the time, when I comment about this, I'm told to take it over and make it work... and to use Claude.
I've kind of decided this is my last job, so when this company folds or fires me, I'm just going to retire to my cabin in the rural Louisiana woods, and my wife will be the breadwinner. I only have a few 10s of thousands left to make that home "free" (pay off the mortgage, add solar and batteries, plant more than just potatoes and tomatoes).
Though, post retirement, I will support my wife's therapy practice, and I have a goal of silly businesses that are just fun to do (until they arent), like my potato/tomato hybrid (actually just a graft) so you can make fries and ketchup from the same plant!
I'm always looking for people to share my weird ideas that have absolutely nothing to do with software or computers. Unfortunately my only friends are all software people who have no interest outside of computers. Something I've found to have very little interest in anymore.
Did you see this last year about the origins of potatoes?
I'd like to get into grafting fruit trees, my uncle was a major fruit eater and filled the yard with many varieties of apples and pears. The apple tree where I live was a mess and I've jsut started the pruning to get it under control.
This might be a little dark, but the majority of our street is very elder, and none of there families want to move over here.
They were the original non-familial homesteaders from 50+ years ago when all this land was my wife's great grandfather's, and he sold off small plots to people. He, infact, inherited it from his father, who bought a half mile square back in the 20s or 30s (I believe). The first house on the road was his (Great Great Grandpa). The road WAS his driveway, then slowly but surely new generations of the family started building houses a few hundred yards away from each other, then they started selling plots to people in the 60s, and sold the last of the original land in 2023 about a year before grandpa passed.
Now the only land left in "the family", is this 1.25 acre plot that I live on. I don't really have the desire to buy more from the folks that are dying, but my neighbor has already bought up about half of the vacant land.
That sounds lovely. I think too many people get attached to the structure of life as they've lived it for the last n years and resist natural phase transitions for far too long. Good luck with retirement and your dream of being the botanical equivalent of the mean kid from Toy Story:p
I noticied what previously would take 30 mins, now takes a week. For example we had a performance issue with a DB, previously I'd just create a GSI (global secondary index), now there is a 37 page document with explanation, mitigation, planning, steps, reviews, risks, deployment plan, obstacles and a bunch of comments, but sure it looks cool and very professional.
Im now out of the workforce and can’t even imagine the complexity of the systems as management and everyone else communicate plans and executions through Claude. It must already be the case that some code based are massive behemoths few devs understand. Is Claude good enough to help maintain and help devs stay on top of the codebase?
I quit my last job because of this. I’m pretty sure manager was using free chatgpt with no regard for context length too, because not only was it verbose it was also close to gibberish. Being asked to review urgently and estimate deadlines got old real fast
Obviously you should also use Claude to consume those 50 pages. It sounds cynical, but it's not. It's practical.
What I've learned in 2 years of heavy LLM use - ChatGPT, Gemini, and Claude, is that the significance is on expressing and then refining goals and plans. The details are noise. The clear goals matter, and the plans are derived from those.
I regularly interrupt my tools to say, "Please document what you just said in ...". And I manage the document organization.
At any point I can start fresh with any AI tool and say, "read x, y, and z documents, and then let's discuss our plans". Although I find that with Gemini, despite saying, "let's discuss", it wants to go build stuff. The stop button is there for a reason.
I use an agents.md file to guide Claude, and I include a prominent line that reads UPDATE THIS FILE WITH NEW LEARNINGS. This is a bit noisy -- I have to edit what is added -- but works well and it serves as ongoing instruction. And as you have pointed out, the document serves as a great base if/when I have to switch tools.
Jump straight to the second option. You have to presume that the content they sent you has no relation whatsoever to their actual understanding of the matter.
We all use Claude at my work and I have a very strict rule for my boss and my team: we don’t say “I asked Claude”. We use it a lot, but I expect my team to own it.
I actually think there’s almost an acceptable workflow here of using LLMs as part of the medium of communication. I’m pretty much fine with someone sending me 500 lines of slop with the stated expectation that I’ll dump it into an LLM on my end and interact with it.
It’s the asymmetric expectations—that one person can spew slop but the other must go full-effort—that for me personally feels disrespectful.
I also don't mind that. Summarized information exchange feels very efficient. But for sure, it seems like a societal expectation is emerging around these tools right now - expect me to put as much effort into consuming data as you did producing it. If you shat out a bunch of data from an LLM, I'm going to use an LLM to consume that data as well. And it's not reasonable for you to expect me to manually parse that data, just as well as I wouldn't expect you to do the same.
However, since people are not going to readily reveal that they used an LLM to produce said output, it seems like the most logical way to do this is just always use an LLM to consume inputs, because there's no easy 100% way to tell whether it was created by an LLM or a human or not anymore.
This kinda risks the broken telephone problem, or when you translate from one language to another and then again to another - context and nuance is always lost.
Just give me the bullet points, it's more efficient anyway. No need to add tons of adjectives and purple prose around it to fluff it up.
Some day someone brilliant will discover the idea of "sharing prompts" to get around this issue. So, instead of sending the clean and summarized LLM output, you'll just send your prompt, and then the recipient can read that, and in response, share their prompt back to the original sender.
I think we'll eventually move away from using these verbose documents, presentations, etc for communication. Just do your work, thinking, solving problems, etc while verbally dumping it all out into LLM sessions as you go. When someone needs to be updated on a particular task or project, there will be a way to give them granular access to those sessions as a sort of partial "brain dump" of yours. They can ask the LLM questions directly, get bullet points, whatever form they prefer the information in.
That way, thinking is communication! That's kind of why I loved math so much - it felt like I could solve a problem and succinctly communicate with the reader at the same time.
If you write 3 bullet points and produce 500-pages of slop why would my AI summarise it back to the original 3 bullet points and not something else entirely?
It won't, and that's the joke. They will write three bullet points, but their AI will only focus on the first two and hallucinate two more to fill out the document. Your AI will ignore them completely and go off on some unrelated tangent based on the of the earlier hallucinations. Anthropic collects a fee from both of you and is the only real winner here.
It's way too early to tell. Safe to say that it's different. But it might be better than some of our current async comms.
If I spend time and thought and research around an idea and a corpus of information and dump that all into an LLM and converse with it, eventually producing an artifact that's partly the LLM's processing of that corpus and partly the result of my direction, and you take that artifact and drop it into an LLM and interrogate it with your own perspective and lenses, that's going to go in directions that I may not have imagined for you but will still contain the kernel of my perspective. And you could indeed interrogate the thing, not just sit back and think about it.
No idea whether this is faster/better or shallower/deeper or if it encourages us to connect more or differently as people or what-have-you. At present I'm not even sure I care, personally, about measuring differences on these traditional axes. It just seems like a vast new communication medium worthy of some exploration so that we can collectively have some idea what we're talking about when we do start to judge it.
> It’s the asymmetric expectations—that one person can spew slop but the other must go full-effort—that for me personally feels disrespectful.
This has always been the case. Have some junior shit out a few thousand lines of code, leave, and leave it for the senior cleanup crew to figure out what the fuck just happened...
Yes, though usually setting up asymmetric expectations requires a power imbalance, so might instead be a PM or someone with influence but not technical acuity creating that initial kLoC.
If you shove content at me that I even suspect was AI generated I will summarily hit the delete button and probably ban you from sending me any form of communication ever again.
It's a breach of trust. I don't care if you're my friend, my boss, a stranger, or my dog - it crosses a line.
I value my time and my attention. I will willingly spend it on humans, but I most certainly won't spend it on your slop when you didn't even feel me worth making a human effort.
I've found in my (admittedly limited) use of LLMs that they're great for writing code if I don't forsee a need to review it myself either, but if I'm going to be editing the code myself later I need to be the one writing it. Also LLMs are bad at design.
I've found that SoTA LLMs sometimes implement / design differently (in the sense that "why didn't I think of that"), and that's always refreshing to see. I may run the same prompt through Gemini, Sonnet, and Codex just to see if they'd come up with some technique I didn't even know to consider.
> don't forsee a need to review it myself either
On the flip side, SoTA LLMs are crazy good at code review and bug fixes. I always use "find and fix business logic errors, edge cases, and api / language misuse" prompt after every substantial commit.
For me it's throwaway scripts and tools. Or tools in general. But only simple tools that it can somewhat one-shot. If I ever need to tweak it, I one-shot another tool. If it works, it's fine. No need to know how it works.
If I'm feeling brave, I let it write functions with very clear and well defined input/output, like a well established algorithm. I know it can one-shot those, or they can be easily tested.
But when doing something that I know will be further developed, maintained, I mainly end up writing it by hand. I used to have the LLM write that kind of code as well, but I found it to be slower in the long run.
Definitely a lot of one-shot scripts for a given environment... I've started using a run/ directory for shell scripts that will do things like spin up a set of containers defined in a compose file.. build and test certain sub-projects, initialize a database, etc.
For the most part, many of them work the first time and just continue to do so to aid a project. I've done similar in terms of scaffolding a test/demo environment around a component that I'm directly focused on... sometimes similar for documentation site(s) for gh pages, etc.
The best thing to do is to schedule meetings with those people to go over the docs with them. Now you force them to eat their own shit and waste their own time the more output they create.
Love the intent, but isn't that wishful if you don't have any leverage? e.g., the higher up will trade you for someone who doesn't cause friction or you waste too much of your own time?
Similarly, managers at my workplace occasionally use LLMs to generate jira tickets (with nonsense implementation details), which has led junior engineers astray, leaving senior engineers to deal with the fallout.
If Claude Code can parse these design documents, I would recommend making a skill to do an adversarial review of the document. Then just generate that review, do some minor edits to make it look like a human wrote it and send it back to them.
Getting similar vibes from freelance clients sending me overly-articulated specs for projects, making it sound like they want sophisticated implementations. Then I ask about it and they actually want like a 30 row table written in a csv. Huge whiplash.
I instituted a simple “share the inputs” along with the outputs rule which prevents people doing exactly this. Your only value contribution is the input and filtering the output but for people with equal filtering skill, there’s no value in the output
The first point is so true. How do people expect me to work with their 20 page "deep research" document that's built by a crappy prompt and they didn't even bother to proofread.
I've definitely seen this, I have a theory as to how this kind of thing actually would affect AI predictions since people seem to only focus on the pure-productivity enhancing effects of AI and discounting the fact that a large portion of work was never productive to begin with...
I've had this experience too. In the case of vibe code, there is at least some incentive from self-preservation that prevents things from getting too out of hand, because engineers know they will be on the hook if they allow Claude to break things. But the penalties for sloppy prose are much lower, so people put out slop tickets/designs/documentation, etc. more freely.
Reminds me of when I was looking for Obsidian note management workflows and every single person who posted about theirs used it to take notes on... note taking workflows.
I like the first blogpost or two. If I recall, it quickly shifts into a pop psychology grindset self improvement book if you keep reading the posts. Its reach starts to exceed its grasp.
>we had to throw things over the fence to IT, DevOps, SRE and whatever name evolutions there were and the reality is AWS/GCE/Azure made things easier than dealing with these teams internally
Anyone who was a dev during the "everyone is devops" fad for a while knows the pain of building something with these kinds of dependencies. Being able to claw back my time from operations on my company's dime is enticing.
There's next to no chance that whatever comes out of the end of this will be a "liberal democratic Iran government". Obama started a route in that direction with the lowered sanctions and the Joint Comprehensive Plan of Action from 2015. Iran having a democratic government doesn't really help the GOP war hawks so of course they trashed it. The same happened with North Korea in the 90s with the Agreed Framework that had some promise before GWB torpedoed it to please his oinking base.
I also think that nuclear powers mean regional stability. Ukraine gave up its nukes in the 90s and we saw what happened there.
He was impeached by the House but that does nothing without the Senate carrying out its trial, which requires an onerous 2/3rds vote. Obviously without the trial in Senate, nothing happens, and nothing ever will until one party gets 2/3rds control.
No it's not. Is there anything to back that up? There's a creative aspect to human research that I've yet to see with gen AI. All it does is regurgitate stuff and get some "new" ideas via the latent space of the distribution it models. But a generative model cannot by definition create anything new. Just estimate its data well enough that it can sample it well enough to fake novelty.
reply