came here from f5 bot - hi I edited this piece! it’s riiight at the tip of outrageous but not totally insane.
which as I look at my own coding evolution over the past few months, is increasingly likely even if I’m not there yet. my main realization is that the SDLC is slowwwly breaking down- first with no human coding and people sending in prompt requests than pull requests. once we figure out the ai replacement for reviews, the software logjam breaks.
> The battery is wearing out a bit, but it started out life with so much runtime that losing a few hours doesn't seem to matter.
this is my exact opposite experience. my M3 Max from 2 years ago now has <2hrs battery life at best. wondering if any experts here can help me figure out what is going on? what should i be expecting?
As others have said, keep the battery in the 80%-30% range. Use the `batt` CLI tool to hard limit your max charge to 80%. Sadly, if you're already down to <2hrs, this might not make sense for you. Also prevent it being exposed to very hot or cold temps (even when not in use)
I type this from an M3 Max 2023 MBP that still has 98% battery health. But admittedly it's only gone through 102 charge cycles in ~2 years.
(use `pmset -g rawbatt` to get cycle count or `system_profiler SPPowerDataType | grep -A3 'Health'` to get health and cycles)
> I type this from an M3 Max 2023 MBP that still has 98% battery health.
That's amazing. I have an early 2023 M2 Max MBP that mostly charges in desktop mode, which limits to 80%. I just looked in battery health and it says 82%. Damn! :(
For giggles, earlier today I asked Apple how much they'd give me for this machine if I traded it in on a brand new $5K M5 Max equivalent. $825. Ouch. I think I will keep it for a few more years. 96GB is enough memory to do anything I want, and it's been such a great performer that it's easily my favorite MacBook ever. I do wish the battery weren't so degraded though.
For anec-science, here goes:
% pmset -g rawbatt
03/03/2026 18:29:51
AC; Not Charging; 76%; Cap=76: FCC=100; Design=6075; Time=1092:15; 0mA; Cycles=63/1000; Location=0;
Polled boot=02/09/2026 07:24:50; Full=03/03/2026 18:24:52; User visible=03/03/2026 18:28:52
% system_profiler SPPowerDataType | grep -A3 'Health'
Health Information:
Cycle Count: 63
Condition: Normal
Maximum Capacity: 82%
Your battery is defective if it's at 82% after 63 charge cycles. My M1 Pro has 87% capacity after ~5 years and 412 cycles of giving zero fucks and regularly draining the battery all the way down to almost 0% and charging back up to 100% every time. I plug in to charge at like 2% super often. Babying the battery doesn't make any sense IMO.
The option to have a 80% cap is being added in the beta versions of MacOS. I think within a few months it should be available to general users without using extra tools.
I set Claude loose on my computer and said “why is my battery life so bad?” and it found an always-running audio subsystem kernel extension (Parrot) which didn’t need to be there and was preventing the CPU from going into low-power states. My battery life got noticeably better when I deleted it.
I’m not even sure how it got installed, possibly when I installed Zoom for an interview once but I don’t know. Point is, at least in one case, AI can help track down battery hogs.
What is your maximum capacity in Settings > Battery Health? What processes are running with significant CPU? What's the typical temperature of the laptop according to a stats app? (Temperature is a good proxy for general energy use.)
I'm typing this on an M3 Max; its max battery capacity is 88%. I've got some things running (laptop average temp is 50-55C, fans off), screen is half brightness, and it's projected to go from 90% to 0% in five hours. I don't usually baby it enough to test this, but 8-10 hours should be achievable.
People here are suggesting limiting your battery charge as a proactive measure to prevent degradation but an M3 is far too new for you to be getting so poor battery life from use, even if you spent all day every day charging and discharging it.
The only plausible answers are either: something you’re running is eating CPU/GPU cycles like crazy (browser tabs gone amok, background processes) or you have a defective battery. Use Activity Monitor to look for energy usage and that will give you a pretty good idea.
I've got a M2 Pro from 3 years ago and battery is still so good I can go to a whole day of meetings and not even need to bring my charger. Then I can probably work all night as well without plugging it in. Battery time is insane.
Unless of course you're doing something that truly sucks down your battery! If I spin up a few Docker instances doing 100% CPU then obviously battery will go down much quicker.
Charge habits with batteries make a huge difference. If your use pattern is that once per day, you take the device from 100% to 10%, you put a lot more wear on the battery than if it kind of hovers in the 30%-80% range for example, or if it just hangs out nearish top-of-charge all day when you're at your desk.
Hot take: people should get used to, and expect to, replace device batteries 1 or 2 times during the device lifetime. They're the main limiting factor on portable device longevity, and engineers make all kinds of design tradeoffs just to make that 1 battery that the device ships with last long enough to not annoy users. If we could get people used to taking their device in for a battery once every couple of years, we could dramatically reduce device waste, and also unlock functionality that's hidden behind battery-preserving mechanisms.
BatFi is a macOS application which will prevent your battery from charging to over 80% by default. macOS does have a version of this built-in but it’s “intelligent charging” I don’t really trust, and I’d rather just have a hard 80% limit except when I override that.
> Charge habits with batteries make a huge difference.
> Hot take: people should get used to, and expect to, replace device batteries 1 or 2 times during the device lifetime.
I agree that people should get used to replacing device batteries, but if you accept that then you should just stop worrying about charge habits. An MBP that doesn't have a defective or extreme-heat-damaged battery should stay above 80% battery capacity for at least 600 charge cycles without any special care at all. That's many years of regular charging, and 80% capacity is still good for all day usage.
Agree, and this is in fact how I treat my devices. I can easily do my own battery replacements on laptops, although I still approach phones with suspicion due to the water ingress problem.
My M3 Max can burn through battery much faster than my M1 Max ever could.
And some apps are really inefficient. New Codex app drains my battery. If you are using Codex I recommend minimizing it, since it’s the UI that uses most power.
A couple weeks ago I was working remote and didn't bring a power adapter, and I realized a couple hours in that my battery was getting kind of low. I clicked on the battery icon and got a list of what was using a lot of power: 1 was an hour long video chat using Google Meet, the other was Claude desktop (which I hadn't used at all that morning).
What in the world is an idle Claude Desktop doing that uses so much power?
just got pinged by f5bot - hi editor here! Personal take: at this point multiple people are already weighing how to remove the human code review bottleneck from agents becoming fully productive. Ankit was brave enough to map out how he sees SDLC being turned on its head and wrote this.
i'm not personally there yet, but I tend to be 3-6 months behind these people and yeah its definitely coming.
It should be a semantic search bot and maybe will be in the future, but for now I rely on the method described at https://news.ycombinator.com/item?id=45546715 and the links back from there.
i vaguely recall about 2 years ago some RAG startup or other did do one for you. i mean this is nothing for a startup and they could use the endorsement ha
Not OP, but having the exact VM spec your agent runs on is useful for testing. I want to make sure my code works perfectly on any ephemeral environments an agent uses for tasks, because otherwise the agent might invent some sort of degenerate build and then review against that. Seen it happen many times on Codex web.
What the other poster here said for testing against a reference, but also as an easier to get started with base for my own coding sandbox with coding agents. Took me quite a while to build one on my own that I was semi-happy with but I'd imagine one solid enough to run cowork on safely might have some deeper thinking and review behind it.
good argument, bad example. GDP is a net revenue number but Stripe is using a gross revenue number (their equivalent of "GMV"). so the numerator/denominator are as different as possible to make it impressive.
reply