How ChatGPT's New Image Generator Stacks Up Against Gemini's Nano Banana Pro
Dec. 17th, 2025 02:30 pm![[syndicated profile]](https://www.dreamwidth.org/img/silk/identity/feed.png){kind=small}
lifehacker_feedFollowing the major image editing upgrades added to Google Gemini back in August—under the whimsical codename Nano Banana—it's OpenAI's turn to supercharge the tools you get for image manipulations inside ChatGPT. The new update is called GPT Image 1.5, and is rolling out now for all users.
One of the key improvements here, as was the case with Nano Banana, is the way that ChatGPT can now edit a specific part of an image while keeping everything else consistent. You can add or remove something, or change the color or style of something, without ending up with an entirely different looking picture.
Another feature ChatGPT has now borrowed from Gemini: the ability to combine multiple images together in one scene. Want you and your best friend in front of Sydney Harbour Bridge? No problem—just supply the source pictures and the AI will do the rest. You can also change visual styles while maintaining consistent details.
OpenAI says the new image editor and generator is able to follow instructions "more reliably," and render pictures up to four times faster than before. Text can be more varied in style and size, and images should be more realistic and error-free in general—though OpenAI also admits there's still room for improvement.
It's the best image generator tool we've ever seen in ChatGPT, and it all looks impressive at first glance—but how does it stack up in practice against Gemini and Nano Banana? I put the two models to the test via the $20-per-month plan on both platforms (that's ChatGPT Plus and Google AI Pro, respectively) to see how they compared.
Rendering and editing images
Open up ChatGPT on the web or on mobile and you'll see there's a new Images tab on the left-hand navigation pane. This takes you to a library of your existing pictures, together with some new prompts for creating images. You get some suggestions for prompts, plus an assortment of preset portrait image styles you can apply.
I tested out the new GPT Image 1.5 model by getting ChatGPT to generate a busy tech journalist, a lamp in the middle of an empty warehouse, and a cartoon-style rolling landscape of hills in the fog. I then got Gemini to create the same pictures with the same prompts. While the results were pretty varied, in terms of quality and realism they were pretty equal—the occasional issue with weird physics and repetition, but nothing too bad.
Both ChatGPT and Gemini are now quite competent at clean image edits, too: Both AI bots seamlessly switched the journalist's clothing to a shirt and tie without touching any other part of the picture. This would have taken a significant amount of time to do manually, even by a Photoshop expert, and shows just how transformative AI imaging is becoming.
Color changes were all handled with aplomb, but the AIs struggled a bit with perspective changes, where I asked to see the same shot from another angle. In these cases, instructions were less well-followed and the images were less consistent (as new areas needed to be rendered), though ChatGPT did a little better than Gemini at getting good results.
The classic "remove an object from this picture" challenge was handled with aplomb: Both Gemini and ChatGPT were able to remove a cottage from the countryside scene with surgical precision, leaving everything else intact. Again, these are the kind of time-intensive image edits that would previously have needed a lot of careful effort, and that can now be done in seconds.
Combining and remixing images
Another talent ChatGPT and Gemini now have is being able to combine images together. So you can have separate photos of you and your parents, put them together in the same shot, and then add in a background of wherever you like. You can get perfect family photos without actually gathering together your relatives together or going anywhere.
This was an area where Gemini and ChatGPT did struggle a bit more: The editing dexterity was still impressive, but the results didn't always look like a single, coherent scene. Lighting is sometimes off, or elements from different images appear at different scales, and you'll have to do a bit more tweaking and editing and reprompting to get everything right.
ChatGPT did fare slightly better at blending different images and elements together, and changing the overall look of a picture. When I tried to get the AIs to mix all my images together in a moody film noir shot, ChatGPT produced something pretty consistent—the Gemini effort looked a lot more like a cut-and-paste job.
It can be fun remixing photos again and again—adding new people, changing the weather, moving the location—and both these bots are now capable of some rather incredible results. Remixing photos of family and friends will be popular, but it's not all that easy: With people you know, any generative AI that gets added tends to look wrong, because neither ChatGPT nor Gemini knows exactly what these people look like, how they smile, how they're built, or how they tend to stand or sit.
In terms of ChatGPT vs. Gemini, they're both at a high level now—a level that puts advanced Photoshop-style editing capabilities at everyone's fingertips. If either AI model has the edge right now, it's ChatGPT's, but there's not much in it. It's also going to be fascinating to see where these image editing capabilities go next.
Princess Jellyfish, volume 1 by Akiko Higashimura
Dec. 17th, 2025 09:11 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
james_davis_nicoll
Can a community of otaku save their apartment building from gentrification? Should a community of otaku save their apartment building from gentrification?
Princess Jellyfish, volume 1 by Akiko Higashimura
These Stylish Marshall Headphones Are Nearly 50% Off Right Now
Dec. 17th, 2025 02:00 pmlifehacker_feedWe may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.
For anyone who’s ever been tempted by Marshall’s signature guitar-amp look in headphone form but couldn’t justify the premium price, the Major V wireless on-ears just hit their lowest price ever, according to price trackers. Originally $169.99, they’re now available on Amazon for $89.99. That’s a 47% drop for a pair of headphones that stand out mostly for their sound signature, battery life, and retro design.
You don’t get active noise cancellation or high-end audio codecs like LDAC, but you do get over 100 hours of battery life (real-life usage may vary), punchy bass, and a retro-inspired design that folds up easily. The headphones run on 40mm dynamic drivers, and the tuning leans toward deep bass and crisp highs. They produce clean, rumbling sub-bass that doesn’t distort even at higher volumes, notes this PCMag review. Vocals stay clear thanks to subtle sculpting in the upper range. It’s not studio-flat audio, but it’s definitely fun. You can further tweak the sound in the Marshall app, which also lets you remap the M button for Spotify Tap or voice assistant access. The joystick on the right earcup handles volume and playback and feels intuitive, something many pricier headphones still haven’t nailed.
That said, there are trade-offs. You don’t get ANC, which means outside noise will creep in, especially with the on-ear design. There’s no Fast Pair, and no pouch to protect them when tossed into a bag. But you do get Bluetooth 5.3, support for wired playback, and surprisingly decent internal mics for calls. Battery life, again, is the standout—well over 100 hours per charge with USB-C or wireless charging via the right earcup. That alone makes the discounted price easier to swallow. If you want headphones that look good, sound bold, and don’t need to block out a plane engine, the Major V makes a strong case, especially while they're under $90.
My Nine Favorite Productivity Methods in One Efficient List
Dec. 17th, 2025 01:30 pmlifehacker_feedWe may earn a commission from links on this page.
A good productivity method can mean the difference between a disorganized, unfulfilling day and one during which you get a lot done and feel great about it. That tradeoff is why so many of these methods, techniques, and hacks exist. Some are over 100 years old, some were accidentally discovered by everyday people looking to improve their lives, and some come from self-styled gurus who publish entire books on their findings.
And some are better than others—though maybe not objectively. To find the one that works best for you, take a look through this guide to seven of my favorites. Notably, these are the ones I like, but there are plenty out there for you to consider, too. Try one that sounds like a strong match for how you think and work (or try to avoid work).
The "Power Hour" method
The idea of a "Power Hour" comes from Adrienne Herbert’s book, Power Hour: How to Focus on Your Goals and Create a Life You Love and asks you to devote an hour a day to working hard on your biggest task—or the thing you care about the most. Sometimes, this is a must-do task that will have major ramifications for your life, like filling out job applications or working on homework. In that case, what you care about most is your overall goal of improvement. Other times, it might be a personal project or passion that you want to carve out time to pursue, which will enhance your life, make you feel happier, and make you more productive that way. The trick is committing—truly committing—to taking that hour every day, upending your schedule if necessary.
Here's my fuller rundown of the approach, which I have started putting into practice in my own life with great results.
The 10-minute rule
Of all the techniques on the list, this might be the one I use (or adapt) most often. Essentially, when there's a small, mundane, or tedious task that doesn't excite you, but won't take more than 10 minutes to do, you just do it. Just do it. That's it. It sounds easy, but it's not, since these are also those tasks you're more likely to put off and ignore, like answering emails or doing the dishes. Fighting against that impulse, committing to just doing these things when you think of them, and then getting them done is hard at first, but becomes a habit over time—and it's one with a lot of benefits. I do this when I'm cleaning my apartment. Instead of laboring over some cleaning checklist or structured plan (which is helpful for a lot of people), I just take on a task when I notice it needs to be done. Then, not only does it simply get done, which is the point, but I feel a sense of motivation and contentment. Since I started doing this, my home has never been cleaner, even though I'm not following any strict guidelines besides "just do it."
I also couple this with another favorite—the "one more" trick, which involves asking yourself, "Can I do one more?" every time you finish a small task. The answer is usually yes and when it becomes no, you give yourself the grace to stop. Just knowing you don't have to, but you can do something can be motivating enough.
Here's a longer explainer on the 10-minute rule.
The “Action Method” of productivity
This method is one of my favorites for keeping on task when I’m juggling multiple projects. It calls on you to organize your tasks into three categories: Action steps, references, and back-burners. Once you’ve done that, you put it all into a spreadsheet with those three categories as the column headers. You slot tasks into each column alongside notes, supplemental material, and whatever else you need—and move them around as they change their designations, as what is a back-burner today might be an action step tomorrow. Organizing it all this way helps you keep on top of the most pressing needs.
Here’s a full explanation of how to employ the Action Method. (The “ABC” method is very similar, with “A” tasks being must-do and high-priority, “B” tasks being should-do activities, and “C” tasks being low-priority ones.)
The 3-3-3 productivity method
Using this technique, you aim to plan your day in threes: Spend your first three hours engaging in deep work on your most important project, then complete three other urgent tasks that require less time, and then do three “maintenance” tasks, like answering emails or scheduling other work. This method works because you do your deep, focused work up-front, which gets you in the zone and gives you a sense of accomplishment and makes tackling the stuff afterward easier. I like this one on days I need a touch of structure. I use a prioritization method—usually MIT, which forces me to think of my responsibilities in terms of the impact they'll have on my life, but sometimes one like the Eisenhower matrix—to figure out which category all of my to-dos belong in, then follow the guidelines to make sure they all get done. I typically pull this one out when I have a big project and am struggling to think of how I'll get it all done. Even breaking one large task down into three hours of hard work, three urgent components, and three "maintenance" tasks helps everything flow a little easier.
Here’s a guide to planning your day in threes.
The “Eat the Frog” productivity method
Similar to 3-3-3, the “Eat the Frog” method invites you to tackle work on your biggest, scariest, wartiest task first thing in the morning. Whatever time-intensive task that has kept you up at night is, that’s what you should do first. It's similar to Power Hour in that you're supposed to do it early in the day, but different in that this may not be your most important or passion-fueled project. After that, everything else should be easier. While most proponents argue you should “eat the frog” as soon as you wake up, this method can work on any schedule as long as you commit to jumping into the hard thing early, enthusiastically, and without hesitation, thus freeing up the rest of your day for other work and lowering your overall stress level. When I use this one, I specify it a little more. If cleaning my kitchen is the most demanding task of the day, that doesn't mean I should do it the second I wake up, but when I get home from work or errands. It wouldn't make sense to postpone leaving for the day to do that, but it does make sense to turn it into the "first thing" I do when I'm in the relevant space, as it makes the rest of my time in the home for the evening better.
Here’s a guide to eating your first frog, so to speak.
The Kanban productivity method
Kanban is similar to the Action Method but requires you to label your tasks as to-do, doing, and done. It works best when managed in a spreadsheet or even on a big board with sticky notes, but you need the three columns so you can move whatever is completed into your “done” pile and anything that still needs doing into “to-do.” If you’re a visual person, this is going to be a game-changer, as it helps you easily see what needs to be done, and gives you some satisfaction when you see what you’ve already accomplished piling up under “done.”
Here’s a guide to implementing the Kanban productivity method.
The timeboxing productivity method
Another trick for the visually inclined and motivated, timeboxing requires you to schedule your entire day. Every activity, from answering emails, to working on a big project, to eating a snack, should go on your calendar. It’s much easier to use a digital calendar, like Google Calendar, for this, since so much of the average day is subject to change and it’s simpler to move things around there than in a physical planner, but try not to deviate from the schedule too much just because deviation is a drag-and-drop away. The idea behind this method is that it allows you to plan to devote exactly as much time to each task as you need to complete it while still filling your entire day with activity. There are a lot of tricks that go along with this one, and once you start, you may go down the productivity technique rabbit hole. For instance, Parkinson's Law dictates that the more time you give yourself to work on something, the longer you'll naturally take—which makes you less productive by reducing the quality of your work over time and stopping you from working on other things. With timeboxing, you can and should practice cutting off time from all your boxes, giving yourself less time to work overall. In the gaps you ultimately create, don't forget to take a break. Those are also essential to productivity.
Here’s a guide to getting started with timeboxing.
The Pomodoro productivity technique
This is an old standby that has withstood the test of time because it works so well: Work for 25 minutes on a task, take a short break of about five minutes, and work for 25 minutes again. Every time you complete four 25-minute cycles, take a longer break. This gets you into the groove of working hard in those 25-minute bursts, since you know a little reprieve is coming. The break recharges you and you get back at it, over and over again, until your job is complete. To maximize the benefits of Pomodoro, get a specialized timer so you don't have to set alarms on your phone and can work without glancing at it and all its distracting apps. On the other hand, an app might be just what you need. I love Focus Pomo, a Pomodoro technique-specific app that blocks all your other ones during "focus sessions." If you're not finding this approach helpful after a few consistent attempts, don't be afraid to adapt it, either. The standard 25 minutes on and five off work well for a lot of people, but you might need shorter or longer work or break sessions. What matters most is that you time them out, commit to taking breaks, and then commit to getting back at it.
Here’s a guide to getting started with the Pomodoro method.
The Results Planning Method (RPM)
This technique comes from famed motivational speaker Tony Robbins, who designed it to be motivational, fast, and efficient. Not only does does the acronym stand for Rapid Planning Method, but it can also serve as a guide to what your day should look like: Results-oriented, purpose-driven, and built around a "massive action plan."
Consistently—every morning or week—ask yourself these three questions:
What do I want?
What is my purpose?
What do I need to do/What is my massive action plan?
By doing this, you connect more to your mission and get more energized about getting to work right away on the answer to the third question, rather than spending a bunch of time deliberating about what you should or shouldn't be doing with your time.
Good Job: My coworker’s emotional support animal is causing me emotional distress!
Dec. 17th, 2025 08:53 am
Adapt This Japanese Productivity Model to Get More Done
Dec. 17th, 2025 01:00 pmlifehacker_feedSo many great productivity hacks come from the Japanese manufacturing industry. The 5S technique helps you keep your workspace clean so you can get more done, kaizen helps you improve the processes and workflow of your job, and kanban helps you schedule your tasks in the most efficient way. These are helpful even if you’re not working within the famed Toyota Production System; they’re adaptable to fit all kinds of work scenarios, which is similarly true for the 3 Ms that originated there, too. The 3 Ms are kinds of waste you should identify and eliminate to keep your work running smoothly. Once you learn how to identify and get rid of them, you can be even more productive, whether you're using kanban, kaizen, or anything else.
What are the 3 Ms?
Lean process thinking, or lean manufacturing, is a production method that seeks to save time within the production process. The 5S technique is en elemental part of lean thinking, as are the 3 Ms. They are "muda," "mura," and "muri," and they all refer to kinds of wasteful practices that slow you down and keep you from being productive.
By identifying and eliminating them, you can streamline everything you need to do. To get them to work in your daily life, you need to adapt them a little, but once you get in the habit of recognizing where things are getting held up and making the appropriate changes, you’ll be breezing through your work.
Identifying and eliminating the 3 Ms
Here’s what each of the three Ms is all about:
Muda translates from Japanese to mean “futility” or “uselessness.” It’s anything that doesn’t add value to the work you’re doing and it comes in two forms: Type 1 is non-value-added activities in your process that are still necessary for your end result, like safety checks, which don’t give the producers of physical goods any kind of financial reward, but do have a benefit for customers and eliminate financial risk for producers. You can’t really avoid Type 1, but you should focus on identifying and eliminating Type 2: Activities that add no value to the process or the end result.
Mura is any kind of unevenness in your operation. Anything that isn’t uniform, regular, or scheduled is mura. Too much mura will result in muda. For instance, if you’re working on a bunch of tasks for a project at your job and your coworker is only working on one or two, it’s actually wasteful because you’ll be waiting around on them to keep moving forward with yours.
Muri is any overburden on a tool or person. It can result from mura, as in the example above, but can also spring from overutilization. For your purposes, think of it referring to you. When you’re overwhelmed and overworked or lack resources, you’re not as productive. That’s muri.
To eliminate muda, identify which unnecessary steps you’re taking in your work. Familiarize yourself with kaizen here, as it’s a process designed to get you to work efficiently. If you find that you’re often doing unnecessary tasks that don’t have much value in the end, like changing the colors and fonts on a presentation deck that’s already completed or constantly tweaking your emails before sending, consider that you might be giving yourself too much time to do them. Parkinson’s Law says that the more time you have to do something, the longer you’ll draw it out and complicate it. Reduce the amount of time you give yourself to complete certain tasks so the pressure of getting it done helps you focus only on the elements that are absolutely necessary.
To eliminate mura, start tracking the process of your work and identify what’s holding you up. Is it waiting on client emails? Is it waiting for a coworker to finish their section of a project? Is it spending too much time in meetings and not enough time actually working? Is it a bad scheduling technique on your end? It might take a few weeks of diligent time tracking, but eventually, the pattern will emerge. If you’re spending too much time waiting on client emails, develop a system whereby you send emails with questions for the next day’s work during the afternoon, giving them time to respond by the time you need to do it, for instance. If it’s your own poor scheduling, start making better to-do lists using kanban or the 1-3-5 method, which reduces your daily tasks to just nine and has you focus on the most essential ones in order of importance.
Finally, to eliminate muri, assess your own burden at work. If you’re missing deadlines, feeling overwhelmed, not being given the right amount of resources, or having too much piled on you, you need to streamline what you’re working on so you don’t completely burn out. Try using the ABCDE method to give a ranking to all of your responsibilities. The A tasks are the most important, followed by the B tasks, but you’ll feel less overwhelmed once you realize the C tasks are optional, D stands for delegate, and E stands for eliminate. Delegate tasks to someone else if you’re overburdened—and don’t do unnecessary ones at all. Some burdens, like meetings, are unavoidable, but when possible, schedule yourself at least one day a week where you have no meetings. Remember to take regular breaks to stay productive, utilizing a method like the time-tested Pomodoro technique. If that method isn't working for you, try the MIT—or most important thing—technique, which helps you reframe your perception of your to-dos around the impact they'll have, not the amount of resources they'll take. Playing around with different approaches is a little time-consuming and might seem like its own kind of waste, but it's necessary to find a system that works for you.
Training yourself to recognize and eliminate these different kinds of waste will streamline your work, reduce your burden, and keep things running more smoothly so you can actually be productive, whether you’re building a Camry or a new investor spreadsheet.
Life lived in dot points
Dec. 17th, 2025 09:17 pmfred_mouseThe damn things continue to overlap
- surgeon appointment: nothing new, but the margins on what was removed aren't big enough, back in surgery - that's my Friday.
- the next step in the candidacy paperwork was in fact not my responsibility, and I now have an email to say I've passed that hurdle (here it is called 'Milestone 1').
- Last Monday rehearsal of the year was this week; I tried bowing for one line of very long/slow notes and ow, nope, not yet. Was, however, good support for the other viola player, including singing some of the bits where the viola has the melody. We had a new violin player! I hope they come back, they seemed to be having fun.
- Today was my last day on campus for the year. I will be working some over the shutdown, because I'm supposed to have my ethics drafted by mid January, and I still don't know what I don't know. Treated myself to curry and a fizzy drink for lunch.
- Finished Building a second brain (Tiago Forte), which I've gained some useful ideas from. Recommended if you are needing a way to organise the information that is coming in to your life; not elsewise.
- Youngest went bouldering with co-workers on Monday, and is learning yet again about not relying on hyperextended elbows to do the work (their grip strength isn't, and their forearms hurt "weirdly")
- have woken up twice this week having done Something Stupid in my sleep. Monday it was the right hip not quite in the right place (went back in during rehearsal, I staggered in looking awful, I gather) and today it is something with the muscles of the right shoulder and halfway down the back -- I could barely move the shoulder this morning, and it has settled down to 'about half the time one or more muscles are spasming'.
Wednesday Reading Meme
Dec. 17th, 2025 08:18 amosprey_archerKate Seredy’s A Tree for Peter, which the library catalog listed as a Christmas book although it has actually just one (admittedly pivotal) Christmas scene. Little Peter lives in Shantytown, a miserable poverty-stricken slum. But his life changes when he meets a tramp, also named Peter, who gives him a red spade and promises to plant a tree for him if he’ll dig a hole for it. Peter does, and on Christmas Eve tramp Peter plants a spruce tree all decorated for Christmas. The candlelight draws the other residents of Shantytown out, and in the warm glow they see that if they worked together to clear out the junk and enlarge Peter’s garden and make the drafty shanties air-tight, they could make this a pleasant place to live… A classic 1930/40s story about common folk banding together to improve their lives.
I also read Ally Carter’s The Most Wonderful Crime of the Year, a romystery that is two part romance to one part mystery which is, unfortunately, the opposite of my preferred mystery-to-romance ratio. I also found it annoying that ( spoilers )
Sadly I think I need to accept that Ally Carter is simply not for me. I’ve tried a bunch of her books and I always come away with the same feeling of “too much boyfriend, not enough spy school and/or mystery-solving.”
By this time I was getting frankly a bit tired of Christmas books, so I took a semi-break with Agatha Christie’s What Mrs. McGillicuddy Saw! (4.50 from Paddington outside the US), which just barely squeaks within the parameters of the Christmas book challenge because What Mrs. McGillicuddy Saw is a murder in a passing train at Christmastime as she is on the way to visit her dear friend Miss Marple.
My first Miss Marple! I’ve been kind of meh on Christie in the past, but I really enjoyed the experience of reading this one although I found the final solution to the mystery somewhat unconvincing. However, I am not reading mysteries for the solution! I read mysteries for the journey and if the journey happens to end in a convincing solution, so much the better.
What I’m Reading Now
This week in Ruth Sawyer’s collection The Long Christmas, a story from the Dolomites about a town of rich, greedy, gluttonous, selfish folk, every single one of whom refused to give shelter to a traveler on a cold Christmas Eve, for which sin the town flooded and became a lake. If you stand on its shores at Christmas Eve, you can still hear the bells ringing for the midnight Mass.
This story is centuries old and therefore not intentionally a parable for global warming and/or the crisis of global economic inequality. However, if the shoe fits…
What I Plan to Read Next
My hold on J. Jefferson Farjeon’s Mystery in White: A Christmas Crime Story has arrived!
untitled 995 (for the road) [by traveller, Joe/Nicky]
Dec. 17th, 2025 12:56 pmsiriareads_feedWednesday Reading Meme & Books 101 & 111 of 2025
Dec. 17th, 2025 07:39 amspikedluvWhat I am Currently Reading: I haven’t technically started it yet, but the next book on my list is Boyfriend Material by Alexis Hall.
What I Plan to Read Next: I have two library books to pick up, so probably one of those.
Book 110 of 2025: The Serpent on the Crown (An Amelia Peabody Mystery) (Elizabeth Peters)
I enjoyed this! ( spoilers )
I liked this book and have already requested the next. Sadly, I think it's the last in the series that doesn't look back at the ‘lost seasons'. I'm giving this one five hearts.
♥♥♥♥♥
Book 111 of 2025: Killing Field (A Jack Reacher Novel) (Lee Child)
I enjoyed this book, but I wasn't sure I was going to. The authors writing style, with all those short, choppy sentences, drove me nuts. ( spoilers )
I liked this book enough to check out the next in the series; I'm giving this book four hearts.
♥♥♥♥
Interesting Links for 17-12-2025
Dec. 17th, 2025 12:00 pmandrewducker- 1. Trans scholars being pushed out of academia, researchers warn
- (tags:transgender bigotry LGBT UK academia OhForFucksSake )
- 2. Scottish Trans and LGBT Youth Scotland response to Scottish Government consultation on updating school buildings regulations
- (tags:school transgender consultation Scotland )
- 3. Magic mushroom compound psilocybin found to break depression spiral
- (tags:depression psilocybin psychedelics )
- 4. Creepy McDonalds AI Actor Responds To Backlash
- (tags:ai advertising actors copyright video )
- 5. The official Fallout season 1 recap is filled with AI and thinks the show is set in the 1950s
- (tags:Amazon AI misinformation )
- 6. Police Scotland warn mental health call-outs are unsustainable
- (tags:police Scotland mentalhealth )
- 7. Tesla Robotaxis crash 12x as often as humans
- (tags:automation driving Tesla )
The Day in Spikedluv (Tuesday, Dec 16)
Dec. 17th, 2025 06:57 amspikedluvI did three loads of laundry, hand-washed dishes, went for several walks with Pip and the dogs, cut up chicken for the dogs' meals, scooped kitty litter, and shaved. I made him cheese sausage for supper (one of the guys at work had it and he thought it looked good).
I watched two more eps of The Pitt. Secrets of the Zoo was my background tv in the evening.
Temps started out at 21.4(F) (it was supposed to be 10, so that was a nice surprise; still cold, though) and reached 31.3. It immediately started going down, but there was no wind (and no more snow clogging the trails) so the walks were actually nice.
Mom Update:
Mom was doing about the same today. ( more back here )
Reading Wednesday
Dec. 17th, 2025 06:50 amsabotabbyCurrently reading: The Magic Mountain by Thomas Mann. Usually in December, after I've hit my Goodreads goal, I read something that's gratuitously long and would otherwise fuck up my goal if it didn't spill over into January (yay for anything and everything in my life being quantified and gamified, love that for me). This year's winner is my high school English teacher's favourite book, which he recommended but said that we wouldn't get until we hit middle age. Well, now I am middle aged so I'm reading it.
It's a curious book. I always hit the literary classics and go like. Oh. Haha. This is stranger and funnier than I imagined.
Me: I guess I will finally read literary classic The Magic Mountain.Thomas Mann: Allow me to introduce my himbo failson, Hans Castorp. He is pure of heart and dumb of ass.
Am I enjoying it? I dunno, as much as you can enjoy a 1000+ page book which goes into detail about the breakfast, second breakfast, rest period, lunch, dinner, second dinner, etc. of the character. Which is the point, really—the mountain in question is a liminal space where in theory, the tuberculous patients can leave, but don't. But it's a slog.
Who's Who in Sea Otter Pup's First Examination?
Dec. 17th, 2025 11:39 amdaily_otter_feed
Last we we showed you a peek inside a sea otter pup admission exam, and here’s a little more info on the process! This comes from the Alaska SeaLife Center, which writes:
Who’s who in a wildlife response patient exam?
When a new patient arrives at the Alaska SeaLife Center, it takes a skilled team to make sure that the first exam is safe and thorough. In EL2526’s admit exam, our veterinary fellow, Dr. Josie, is the one performing the medical assessment, while Wildlife Rescue and Animal Care Specialist Savannah is handling the pup.
The handler plays a critical role. Savannah’s focus is on protecting both the veterinarian and the wild animal by staying in just the right position to prevent potential bites while also not stressing or applying too much pressure to the animal. You may also notice the gear. Savannah is in heavy bite gloves, and Dr. Josie is in nitrile exam gloves.
1 week until deadline!
Dec. 17th, 2025 10:31 amchacusha posting in ![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
latetreatbonanzaBe sure that your works are ready to go live by then. We currently have 5 late treats in the collection!
Sign-ups are also still open. It's a bit last-minute to sign up, but you can go ahead and do it before LTB ends for the year!
Useful links:
Schedule & sign-up post
Late Treat Bonanza 2025 AO3 Collection
Searchable sign-up spreadsheet
Event rules & FAQ
Just One Thing (17 December 2025)
Dec. 17th, 2025 09:12 amnanila posting in awesomeersComment with Just One Thing you've accomplished in the last 24 hours or so. It doesn't have to be a hard thing, or even a thing that you think is particularly awesome. Just a thing that you did.
Feel free to share more than one thing if you're feeling particularly accomplished! Extra credit: find someone in the comments and give them props for what they achieved!
Nothing is too big, too small, too strange or too cryptic. And in case you'd rather do this in private, anonymous comments are screened. I will only unscreen if you ask me to.
Go!
Write every day: Day 16
Dec. 16th, 2025 09:14 pmluzulaTally:
( Read more... )
Day 15:
sylvanwitch, trobadora, brithistorian, badly_knitted, goddess47, the_siobhan, cornerofmadness, carenejeans, china_shop, sanguinity, chestnut_podDay 16:
china_shop, brithistorian, sanguinityBonus farm news: Housemate is sawing up a section of a large oak trunk that we got when the neighbors had a tree cut down. Among other things, he plans to make a large oak table for the living room out of two of the (very wide and heavy!) planks. Which will now need drying for two years before said table can be made, but I think it will be gorgeous.
Fandom-dropping update: A secret fourth strategy
Dec. 17th, 2025 02:33 amerinptahIn that last post about pruning my tag-wrangling assignments, I mentioned the “got 1 fic, the fandom tag got canonized, then nobody ever used it again” fandoms.
Then it occurred to me that there’s an even smaller type: the “nobody used it again, and the original author deleted, so the fandom tag is still around but has 0 works” fandoms.
If there’s an automatic way to find how many of these you wrangle, I don’t know what it is. So I just did a pass through my fandoms with 0 unfilterable/unwrangleable tags attached, and double-checked the work counts.
You know how AO3 will give you a “Retry later” error if you try to load too many pages in a row? Yeah, it made me take at least four breaks while I was going through this process.
The payoff is, now I’m down to 1338 fandoms. Knocked a full 62 empty tags off the list.
(Their tags are still canonical, but they won’t show up on the Unassigned Fandoms list, until/unless some user posts another fanwork that makes them 1-use again.)
December Days 02025 #16: Badger
Dec. 16th, 2025 11:17 pmsilveradept16: Badger
If you like, put on Badger Badger Badger. (Personally, I'm here for the Mosesondope.EXE Demoscene Edition, because of the visuals and the sound, but the original is also good and will serve.)
The key word for this entry comes from the Lucy A. Snyder piece Installing Linux on a Dead Badger: User's Notes, and then the subsequent book, Installing Linux on a Dead Badger (and other oddities) that took the single article and created more from that world, packaged up into a small set of pages.
I'll admit that replacing the operating system on a computer is not a task for the faint of heart. This is also another one of those cases where having a spare machine, or even a spare hard drive to devote to the installation, makes things much more approachable than they might otherwise be in trying to make things happen with just one drive and one computer and therefore one chance to get it right, without additional help from recovery utilities. For a good amount of time, I had Linux on one drive and Windows on another. Actually, I still do, it's just that Linux now gets the faster and better drive now, instead of Windows.
To make some of those situations less frightening, there exist things like the Windows Subsystem for Linux, that you can install and enable and then download a compatible distribution to get terminal access to that distribution. (It doesn't do GUI.) Or there were projects that basically set up an image inside the already partitioned drive, so that there wasn't any need to repartition the drive and worry about what might happen to data as it gets shuffled around. And most Linux distributions have a live environment on their install images so that someone can at least poke around a little bit and see what it might be like to use that particular distribution, how it manages software, what things it includes as a default, and what desktop environment it believes is the best one, and therefore is the one it puts forward as the default option.
Because just about every Linux distribution is Opinionated about these things, it can take a certain amount of trial-and-discard before you find one that you're willing to work with long enough to figure out whether you're truly compatible, or whether it still does things over time that will annoy you to the point where you find you have developed Opinions of your own about how you want your Linux to function, and then you will be able to read distribution documentation and hype statements to find out whether or not their Opinions match yours. I started trying to run Linux in my undergraduate university days, and the experience was so rough I jettisoned the idea entirely for my undergraduate period. By the time graduate school rolled around, and a Linux environment was the best option for me to do some of my schoolwork, instead of trying to flatter Apple by buying into OS X, sufficient improvement had happened that the process of installation and use was much more on the Just Works side instead of the problem-ridden situation I ran into. It could be that I selected very poorly in my first distribution choice, as well. In any case, graduate school and my first few years of independent living were relatively smooth in terms of making it all work out, and I had a Linux machine with a TV tuner that I could use to watch most of the cable channels I had available to me. in my apartment.
So, in the sense of not necessarily needing large amounts of technical skill and fiddling with configurations, and that most distributions contain an installation wizard to set specific environment variables, partition drives appropriately, and install the software, it's not that difficult to install Linux on a desktop machine. Laptops are a little fiddlier, and the single-board computers, like the Fruit Pi lines, are the fiddliest of the lot. That said, the desktop installers work 99% of the time for laptops, and the Fruit Pis generally have their own installers / image writer programs to ensure that everything gets put in the right place. Significant amounts of work goes into the installers to make sure that they function well, cleanly, and without errors, so that someone can feel confident that the potentially most dangerous part of the process is simple enough, and that all of the options they need to make sure their machine will come out the other side running optimally and with any and all of the tweaks or packages it needs to do so already enabled.
The story that provides the keyword for the entry is much more like what it can be to try and port Linux to a new set of hardware, or trying to figure out how to get all the drivers in place and running smoothly, and any adjustments that need to be made to the kernel to make it happen. All of which is arcane wizardry well beyond my level of current understanding. I am in user space, not in kernel hacking space. (And there you can see why I think "Oh, I'm just running other people's software" is an appropriate deflection for any kind of praise for things that I'm doing with that software.) It's a lot easier than it has been to install Linux on a dead badger, or any other animal of your choice, than it has been before, and it will likely get easier as time goes on and the installers and distributions are refined even further, and Linux is available for a wider range of possible hardware and components attached to that hardware. Because Linux people want us to adopt a distribution (and preferably theirs), they're trying to make it as simple as they can to get it done. So having done it several times at this point, and changed distributions, and mostly just used the tools available to do it with, I don't consider having installed a Linux to be a particularly praiseworthy thing for me in most circumstances. (It's recipe usage. Just follow the directions and you'll be fine, pay no attention at all as to how following recipe almost always has an underlying assumption that you know all the techniques that it's going to ask you to do.)
George, the original-model Chromebook, is an exception to this. It was still recipe-following, but I had to be a proper information professional to find and extract the recipe from where it was being stored. Pulling the same feat again with a different model of Chromebook is pretty impressive, since that still meant finding the appropriate spots on the circuit board to disable the write protection and doing the thing the recipe needed to do that disabling. (I think it was removing a screw, in this case, instead of using electrical tape to prevent a connection.)
Putting aftermarket operating systems on phones and tablets is still recipe following, but in my case, for Android things, it requires operating from the terminal to achieve the desired results, as well as manipulation of buttons or finding ways to ensure that the correct places are being booted into to use those recipe tools. And while I've had success at every item I've attempted, there was one time where I straight-up botched the process by flashing the wrong thing to the device! While this would normally be a straight-up brick problem, on this specific device (an old Amazon Kindle Fire), with some digging in the information and reading more of the troubleshooting parts of the recipe, it turns out there's a pad on the circuit board that if you create the right kind of short to it, you can force the device into a firmware-upload acceptance state for a little bit of time. Which involved the dexterity and care needed to disassemble the device to expose the pad, to have the right kind of wire on hand to create the short, and to have the terminal command only needed the carriage return, so that I could hit my window of opportunity and flash the correct item to the device. And then, after that, to reassemble the device, after confirming that it had, in fact, taken the correct flash and could now function properly again. That was an adventure, and it'll teach me to read things more carefully the next time I get a wild hare in my bonnet about doing various things. (That said, this device was old, it was not mission-critical, and while it was much improved for having been put on this path, it still wasn't a very powerful device, and the version of Android built for it was several version numbers ago. So botching the flash was a question of whether I could do the recovery, not whether I had to do the recovery. Much less pressure.)
[Diversion: There is at last one cellular device carrier who locks the bootloaders of all their devices and refuses to provide any means of unlocking them to their consumers. The problem is, unless the seller already knows, and/or has already installed an aftermarket operating system on the device, there's no way of knowing whether a used phone that you're interested in will be one that you can put aftermarket software on, or whether it will be one of these bootloader-locked devices from this carrier. It's remarkably hard to source new old devices because of this, and there's enough confusion between carrier unlocking, where a device can be used on any of the carrier networks in a country, and bootloader unlocking, to install software, that a device proclaiming itself "unlocked" is often carrier-unlocked, and unknown about whether it's bootloader-unlocked. I would happily source a device from the manufacturer to avoid such nonsense if I could, but buying direct from the manufacturer is often hella expensive, and needs to be paid all at once. (That, and they usually only carry their newest models of the niches they're looking for, so trying to sneak an older model from them usually is a no-go.) I'd rather test phones that I'm getting from the used market for their suitability before buying them, but that can also be difficult to do over the Internet, unless, of course, the seller knows what I'm asking for and can do those tests themselves and show me the results.]
There are two exceptions that I know of to the idea of making Linux easy to install and then just use, so long as you agree with the opinions of the distribution creators. The first is Gentoo, which, having now read the Wikipedia article on the distribution, seems to have a few more options for providing pre-built ways into a system, that then get taken over by the way that a Gentoo system really wants to work, and was previously installed: by compiling everything from source, according to preferences set by the user to ensure that the software that they used was exactly the way they wanted it to be (and that had been optimized for their hardware and use cases, so that it would go faster and potentially use less RAM on that system compared to others that had not been optimized.) Even I, supposed computer-toucher and polymath-in-training, have never attempted to stand up a Gentoo system according to the official instructions and handbook there. Just from how it is described, I feel like it offers a jam choice problem to everyone who doesn't already know all the answers to the questions it will ask in advance, and therefore can just set the flags and switches in the manner they desire and leave the machine to compile everything. (Plus, updating the system for Gentoo has to take significant amounts of time to do all the compiling, so I would hope that the performance improvements more than make up for the increased amount of time spent building all the packages from source.)
I have, however, tangled with the second exception to the rule, and stood up several Arch Linux systems using their official methods. Arch's Opinion on Linux is that they actively try to avoid having one, past making sure that packages are built according to their specifications, and that they do not particularly care for large amounts of abstraction. Past the basics to get a system up and running, they have no defaults, they have no recommended packages, they have no application suite or desktop environment that they install by default. There are now a couple ways to go about setting up an Arch Linux system, one with a guided installation and one that follows the official installation process on the Arch Linux wiki. The Arch Linux wiki is the reason that Arch Linux isn't a niche distribution that only the hardest of hardcore users takes on. The documentation on the wiki is excellent, although sometimes it is esoteric, and the documentation is frequently more helpful than the forum users, who often demonstrate the kinds of stereotypical attitudes that people have come to think of Linux users, and of people who generally are unhelpful until you do the exact thing they're demanding, at which point they may give you a curt answer with no explanation to help you understand. So, for someone who believes in their ability to follow recipe, having a nice detailed recipe to follow and to refer to when things get a little squirrely is just the thing desired.
Thankfully, despite the flaws of its users, Arch is a distribution that fits with the idea of how I wanted to work with systems. On other systems and architectures, the developers and maintainers make easy the pathways they want users to use, and make very difficult pathways that the user might want to take that aren't what the maintainers want. And the update schedule for many distributions is slower than what I would like it to be. Debian updates, but there's enough that's been changed in the interim that I wouldn't be surprised if they recommend reinstalling rather than version upgrading in place. Ubuntu releases every six months, and Mint follows that schedule. Arch (and Gentoo) and their derivatives are, instead, rolling-release distributions, where updates to the packages are available immediately, rather than at specified update intervals. By remembering to keep the system updated regularly, or at your own decision of intervals, you can control a rolling-release for your own schedule, rather than having to set aside time when someone else wants to update. And because an Arch install from something other than the guided script has very few decisions made for you, it's perfect for cobbling together all of your favorite programs in one distribution, never mind how they might have clashing appearances with each other, based on what toolkits they're using for graphical styling.
Valve Corporation, in creating the operating system for their Steam Deck devices, SteamOS, took Arch Linux as the base and provided significant support to the distribution to ensure that it could continue to be used as the base for SteamOS. And, as they have done with many other things, the corporation created a very nice wrapper around Arch so that they could run Steam in Big Picture Mode on the device, and still allow for people to use the Deck as a desktop-style device, or to game in high resolution and power if they hooked it up to a dock. Arch's aggressive opinions about keeping the decisions in the hands of the user make it a great base for Valve to build upon, since they can choose what they want to apply and only what they want to apply.
All of the pure Arch installs I've done so far have eventually wound up with certain kinds of problems that necessitated their reinstall or my choice to move a machine to another system. Usually it had to do with storage problems. On the one that had enough storage, the problem was essentially that the discrete graphics card in the machine was no longer supported by the proprietary drivers for the corporation that made it, and so the desktop environment choices were very limited, and even then, the machine was starting to struggle with doing all that many things. These could have been defeated in various clever ways, but eventually it was clear that the problem was going to only keep creeping up on me and getting thrown back after a little bit. So, admittedly, having done the thing, I eventually abandoned doing the thing for a more opinionated setup that still runs the Arch base, but goes from there to provide some better quality-of-life features, and which has a gaming edition, which is what I wanted in the first place. I'm not sorry that I did the Arch from scratch approach, it introduced me to some neat tools that I can use in the future if I ever need to stand things up, or use console commands to try and achieve various situations like starting up wireless Internet. And, doing the whole thing from the command line meant that I got a little more confident in my abiity to do things from the command line. (And, subsequently, understand why there are so many warnings on the Internet about not using combinations where you download code you haven't looked at and then pipe it directly into your shell. Even though I probably wouldn't be able to examine such scripts to see if they were malicious before executing them. I don't have that kind of specialized knowledge, I operate firmly in user space, and so I do a lot of trusting that the people providing this software aren't doing it for malicious reasons, and that they haven't had someone introduce malicious code into their project, whether by force or by social engineering to get themselves attached to the project and then push malicious changes. That it's worked out marvelously so far is a testament to what people can do when they cooperate with each other and are able to use tools at their disposal to sign their work and make sure that it's trusted.
Installing Arch isn't quite like installing Linux on a dead badger, that would have been if I'd managed to successfully get the Linux experiment I did in my undergraduate days to get up and running and doing what I wanted without a lot of frustration and aggravation. But it is something that rightly suggests that at least my abilities to follow recipe and to troubleshoot when something other than the expected result comes out of the recipe, so as to get it back on track and working again. This happens regularly, and in the distributions that I'm running now, the updater script they provide tells me when there are things that may need my attention, like new configuration files that I might have to tweak to make work again. It's good, it's powerful, I like the aesthetic of it, and the machines that need to run it can. (And I'm still taking care of all the rest of the fleet of devices as well, with their specific purposes in mind. Update day is usually an all-day affair for all of my items, but the nice thing about package managers and update scripts is that they do most of the work for me, and I just have to run the commands to make sure everything is in order.
So, if there aren't reasons why you have to stay on a particular operating system, maybe give a Linux a spin for a bit and see if you like it? Or several of the Linux-type items for a spin, and see if any of them appeal. Each time Microsoft decides a Windows version gets no further security updates, or Apple decides that certain computers no longer get macOS updates, or phone manufacturers decide they're done providing updates to their devices, there's an opportunity for a Linux to step in and keep the device going, so long as you can install it. And so long as you trust the community of developers who are interested in keeping that device going past the end of official support from the manufacturer. It's worked out for me pretty well since I switched to Linux as a primary driver of things, and now I can say that a lot of gaming is actually coming along nicely on Linux systems, so that's pretty cool.
Photos: Testing Pens on Plant Labels
Dec. 17th, 2025 12:42 amysabetwordsmith posting in journalsandplannersThese are the other posts regarding the labels.
1/3/25 Photos: Testing Pens on Plant Labels
2/3/25 Photos: House Yard and South Lot
3/3/25 Photos: House Yard and South Lot
4/4/25 Photos: South Lot
5/6/25 Photos: South Lot
6/2/25 Photos: House Yard
11/3/25 Photos: Lantern Terrarium Assembly Part 2 Testing the Fit (labels at bottom)
Photos: House Yard 12-16-25
( Let's do science to it... )
Sang in a concert
Dec. 16th, 2025 10:18 pmsoniaI had a small trio part in a Serbian song, and then a solo verse in a Ukrainian song where there were 17 (!) short verses and we each had one, except the last one we all sang together.
It all came together! I was nervous, but it all flowed, and I'm getting better at being able to open up and sing even with an audience there. As the sessions go by and we all get to know each other and get more comfortable with performing, the ambient nerves settle down and I have an easier time managing my own nerves. I used to outright panic, and now I worry a fair amount beforehand, but by the time the concert itself rolls around, I figure I'm as prepared as I'm going to get.
So grateful to get to sing with this teacher and these singers every week. This is a big piece of what I came back to the Bay Area for.
Sometimes I finish things
Dec. 17th, 2025 04:53 pmmergatrude( cut for photo and blather )