Agents are great (in so far the models are able to complete the task). Autocomplete copilot just feels like bad UX. It's both, not super effective and also disruptive to my thinking.
I think it depends on the context. If I've been writing the same language and frameworks and code solidly for a few months, then autocomplete gets in the way. But that rarely happens, I like to keep trying and learning new things.
If I'm familiar with something (or have been) but not done it in a while, 1 - 2 line autocomplete saves so much time doing little syntax and reference lookups. Same if I'm at that stage of learning a language or framework where I get the high level concepts, principals, usescases and such, but I just haven't learned all the keywords and syntax structures fluently yet. In those situations, speedy 1 - 2 line AI autocomplete probably doubles the amount of code I output.
Agents is how you get the problems discussed in this thread: code that looks okay on the surface, but falls apart on deeper review, whereas 1 - 2 line autocomplete forces every other line or 2 to be intentional.
Besides being trivial to enumerate, every service and their mother is asking for a phone number, from physical stores/rentals/hotels who sometimes reject you if you say no, to the countless online services either asking for it or requiring it. I must have given away my phone number to hundreds if not thousands of entities after having it for more than 20 years.
> None of this is particularly difficult technically. Even simply slapping x509 certs on calls and having some basic filtering would achieve a lot.
Slapping x509 certs on probably some of the oldest telecommunications infrastructure in the world (both in terms of devices using it, and devices enabling it) wouldn't be "technically difficult"?
But I've never worked in telecommunications, maybe I'm overestimating how large piece of work this would be.
I think you got your proportions slightly wrong there. This will be contributing as much to an AI bubble as a kid tinkering around with combustion is contribution to global warming.
Not really. Anything that guy does sets the tone for an extended cacophony of fans and followers. It would be a sad day when nobody critically assesses the motivations, effects and framing of those moves. I question the claim this move helps humanity and stand by the assessment it's just more feeding an unfree ecosystem which equates to propping up the bubble.
A year ago I was using GitHub Copilot autocomplete in VS Code and occasionally asking ChatGPT or Claude to help write me a short function or two.
Today I have Claude Code and Codex CLI and Codex Web running, often in parallel, hunting down and resolving bugs and proposing system designs and collaborating with me on detailed specs and then turning those specs into working code with passing tests.
The cognitive overhead today is far higher than it was a year ago.
Interesting. To me it does not really make sense to think in terms of fuel left because, no matter the reserves, there can always be a situation so unlikely, so outside the ordinary, that it will drain all fuel reserves before you make it to the planned destination.
The moment I have been waiting for: Top 1% player in all scenarios, both Mini Metro and Motorways. AMA.
No, they are both really fun (and highly addictive in my case). I like that you can do a scenario in 30-ish minutes (and even pause if you need to). I personally prefer Motorways over Metro, but alas, both highly recommended. Fantastic game design.
Thanks! I've played the daily Mini Metro challenge for years and very rarely made it into the top 10%. I've gone through phases with widening loops, grids, etc. I always feel like there must be some mysterious "trick" I don't know. Questions:
1. How important is it to make sure you alternate symbols? (Beyond the obvious of not having 3 in a row). Do you go out of your way to avoid two in a row?
2. Is it better to put major junctions at the most common circle/triangle symbols, or on squares, or the rarer ones?
3. How much imortance do you place on the slowdown from lines crossing not at stations? I always go out of my way to avoid doing it but I wonder if I overrate the importance of it in my mind.
4. When you notice that some random station along a single line is getting a lot more traffic than other ones, do you shift other lines to cross it or just add more carriages?
One of the most frustrating (but addicting) things with the game is that a couple of my highest scores happened when I first started playing it, before I thought I knew any tricks at all! Wish I could see what the best players' maps end up looking like.
Oh yeah, one more question... do you play the secret level? What actually happens there, or is it just a gag?
1. Alternation matters because circles spawn most often, triangles next, squares least. I avoid 3+ in a row almost religiously; 2 in a row is fine unless that segment is already stressed. If I can cheaply flip a segment into triangle-circle-triangle (or similar), I do it.
2. Where to put big junctions: Squares are rare magnets, so giving every line access to at least one square prevents square-bound riders from piling up on transfers. My "major" hub is usually a square near the geographic center or a bridge choke, and I try not to merge too many trunks into one block-split load across two nearby interchanges if possible. If you have special shapes (stars, etc.), connecting each special to exactly one line that meets others only at an interchange keeps flows predictable.
3. Crossing lines away from stations: It's not catastrophic, but on busy trunks the cumulative hit adds up. I'll avoid mid-block crossings on high-throughput segments; elsewhere I don't contort the whole map just to eliminate a single crossing. Net: treat it as a tax, pay it only when it saves tunnels or ugly detours.
4. Random station getting swamped: Triage in this order: turn it into an interchange (boarding speed), add a loco, then a carriage on the most burdened line, and only then re-route another line to cross it. If nothing stabilizes, drop a short "shuttle" micro-line from that station straight to a square/triangle sink, then delete it once the queue clears. Pausing to redraw aggressively is part of the game.
On those early "beginner's luck" highs: seed luck is real but also the game rewards ruthless mid-game rebuilds more than early cleverness. Don't be sentimental about lines-pause, re-lay, and keep every line touching a square.
Thanks for the tips! I haven't normally redrawn as aggressively as you suggest. I'm excited to try all these strategies in practice, and it's enlightening as well to hear your thought process.
The secret level plays in endless mode. On the home screen of the game, tap the white arrow in the upper left corner to go to the credits page. Then return to the home screen and go to the credits page again. The second time there will be a sort of gear looking design next to the credits. Tap on that.
I don't know if there's a further easter egg to solving it but maybe you'll figure it out ;)
I have a similar set of questions about a different puzzle game.
I've reached a point where I don't feel I'm improving and I'm not discovering any more tricks.
Going online to find other people's solutions just feels wrong. Like buying a "how to complete $X" from the olden days. I don't learn when someone else provides the way to a solution; I like progressive learning, and I dislike the frustration when I stop progressing.
I keep wondering about the meta-problem of how to change my mental gears...
There's an itch to create my own version of the game which would force me to learn the underlying mathematics. Most games are designed to level your skills or knowledge up slowly (sometimes a social component involved). Building yourself forces one to understand the constraints better.
OR do I just train up my meta-skills e.g. playing similar but different games?
I do feel stupid!
Obviously I've also reached my limit on the metagame here, given that I'm asking...
When you reach an impass with a solo intellectual challenge, what's the next step?
Well, I don't usually ask for help and I don't want solutions handed to me, but getting another player's thought process doesn't seem like cheating. I suppose it depends on whether there's one trick to always solving the puzzle, in which case learning the solution without solving it yourself ruins the fun, or whether it's a game that you can always keep improving at and never completely be perfect at (like Mini Metro, which everyone loses eventually, or Poker, or chess, etc) where sometimes learning from others' strategies helps you level up and become familiar with concepts that unlock further skills and open up new questions as well.
Wow. AMA, you say? When do you prefer loops, or lines? Additional lines, or carriages? How much do you tear down at once? if you don't mind me asking
I'm like median on Metro, ~60 hours over years (though perhaps just the one hour, 60x, &c). Never too late to learn some strategy, I guess. Never played Motorways.
Loops vs lines: Loops only in a dense core where you can keep shapes alternating and include a square. Use lines for suburb<->core and river hops.
Lines or carriages: Early add lines, midgame add a loco, then carriages on the trunk that is actually redlining. Late add an interchange at the first overloaded transfer before more cars.
Tear-down: Hm... how much, not sure how to quantify. Definitely something you must do in every long running game but the extend is different. As a heuristic: Pause and rebuild when queues outrun a single weekly upgrade. Reorder shapes, make sure every line touches a square, split any mega-hub into two nearby transfers.
As you probably have guessed: There is no real silver bullet. Knowing the best move is basically impossible, the space is too complex. As a most useful general skill, it's important to recognize problems very early and optimize ruthlessly.
> How do you balance the struggle to recognize your own greatness while also making time to engage with the little people?
Alas! I lie awake many a nights.
> Also, how much time would you say it's taken you to refine your skill to get to that 1%?
Just checked Steam stats. Surprisingly (well, to me) little: Around 60 hours of Mini Metro and 200 hours of Mini Motorways. I guess it's not exactly competitive esports.
reply