First Impressions of the Pixel 9 Pro

One of my technological indulgences is that I upgrade my phone yearly, and for the last several years this means I get whatever the newest iteration of the Pixel phone from Google is. This year it’s the Pixel 9, and I opted to go with the “Pro” version. It has slightly improved cameras plus a bunch of “AI” integration on the phone and on the cameras, and I was curious to see what that meant in the real world. I got it yesterday, took a bunch of photos and played with some of the new features, and here is a (very) preliminary set of thoughts about it.

1. This year Google split up its “Pro” line into two phones: The Pro and the Pro XL. In terms of technical and camera specs, these two are the same phone except in three areas: The size of the phone, the resolution of the screen, and the battery capacity. In all respects the “Pro” is smaller; it has a 6.3″ screen, compared to the 6.8″ of the XL, and has fewer pixels on the screen (although due to the size difference it has higher pixel density), and a slightly less than 5000 milliamp battery.

I got the smaller version because while I liked my Pixel 8 Pro, its 6.7″ inch screen and overall corresponding size was just a smidge too unwieldly for me. The smaller size fits me better; it’s rather more easily one-handable. The 9 Pro still not small — my favorite version of the Pixel, sizewise, was and likely always will be the Pixel 5, with its modest 6-inch screen — but it is manageable, and I appreciate that.

Internally the thing packs 16GB of memory, and I got mine with 512GB of storage, the former to handle the “AI” toys the phone has and the latter because, well, I take a lot of pictures. I got the “Hazel” color, which is a pleasant gray-green, and the back and front are both glass, which means, for me, anyway, it’s slippery as hell. I put a clear case on it as soon as I transferred everything over from the Pixel 8 Pro.

The takeaway here is that I like the way it looks and feels, and while I don’t expect Google to keep the smaller size for the Pixel Pro 10 — smaller-sized phones almost never seem to be popular — I’m glad to have it for this iteration.

2. Google has kept the camera hardware mostly the same from last year, with the exception of upgrading the selfie camera to 48MP, and tweaking the ultrawide camera slightly. Most of the improvements, as they were, come from the software. By and large, I think this is fine. I am a big fan of the Pixel cameras (it’s a major reason I keep getting the phones), and a quick set of photos suggest to me the things I am most happy with about them continue to be the same. Pictures look great right out of the camera. Here’s a quick selection of some of the photos I took yesterday:

These last two photos were taken at 30x, which means 5x optical zoom with Google processing. The moon shot is tolerable but nothing on what I could get out of my Nikon in terms of clarity. But then, I don’t expect the Pixel to outcompete a big sensor and big glass on things like moon shots. I do expect it to be decent in everyday shots, and it does a very good job of that.

I will note that Google has improved the “panorama” function, which makes it easier to use, and it does a better job of stitching the pictures together, although it’s still not perfect:

The Pixel portrait mode still struggles with fake bokeh, so I recommend just taking your portraits at 2X zoom, it’ll work just fine.

In addition to the previous set of computational photography tools, Google has added a “magic” editor that will let you use “AI” to generate new images on top of the pictures you’ve taken. I tried it a couple of times. Here’s the first one, where I asked it to replace the sky outside my house with a mushroom cloud:

And another where I asked it to place Krissy in an open field with rainbows and kittens:

Meh. The mushroom cloud is fine, but the rainbows and kittens picture just looks like a bad Olan Mills backdrop. As a general rule, I’m not a huge fan of “AI”-generated work because it largely looks like hell and is also a copyright nightmare, so I don’t expect to see myself using this much in any event, except possibly to erase backgrounds (the Magic Editor does do a pretty decent job at edge detection).

Google has another new photo mode in which you can take a picture of a group, and then have someone else take another photo where you line yourself up with where the group was, and put yourself into a computationally spliced photo. I haven’t tried it yet, but I’ve seen others try it, and I’m wondering what it offers that mostly can’t be replicated with the front-facing camera on ultra-wide setting, held in landscape mode. I’ll have to try it out for myself and see what I think.

In general, however, it seems like the photo “AI” and other improvements are mostly gimmicky, which is fine, but not something I’ll use much. I’ll just let the Pixel do what it’s done for a while now — take really good pictures — and then if I have any editing to beyond basic color balance, I’ll shoot the pics over to Photoshop.

I’ll be taking the Pixel 9 Pro with me to Brazil later this week and we’ll see how it does there. I expect it will do just fine.

3. What about the other “AI” stuff in the Pixel phone? Well, Google now stuffs Gemini into the Pixel, and I believe the eventual plan is to have it replace Google Assistant. But inasmuch as Gemini is up to its usual bullshit, I’m not in a rush to use it all that much:

I’ve mentioned before how fucking cowardly I think Google is, first in obscuring information about basically anything relating to US politics in their Gemini LLM, which it otherwise purports to be a useful replacement (or at least, a front end) for their search, and secondly by insisting on using an LLM front end that it knows hallucinates misinformation and is easily manipulated into offering up that misinformation, to such an extent it cannot be relied upon to answer basic questions like “Who are the current candidates for president in the United States” and “Who won the 2020 US presidential election.”

Anyway, I’m not planning on using Gemini for much of anything if I can avoid it, it’s not trustworthy. Also, fuck you, Google.

Google also offers Pixel Studio, which is an image-generating program; you put in a prompt and a likely copyright violation pops out. The images (which do not allow humans in them yet, I expect that functionality to come online after the US election) are fine in that uncanny way most “AI” images are at this point:

Be that as it may, I am not the audience for this functionality, since aside from showing the above as an illustrative example, I get all my art from actual humans and recommend you do too. Making “AI”-generated bullshit to text pictures of frolicking mammals to your mom (or something) is a venial sin, I would say, but this nonsense is showing up in memes and book covers and etc, and it’s already exhausting to deal with. Just let it be, y’all.

4. In fact, I would say the best things about this new Pixel 9 Pro are the things not heavily marketed as “AI” anything, just solid improvements to an already very excellent phone (and it is, also, a very excellent phone — the phone calls I’ve gotten on it have been perfectly clear, and the suite of phone functionality Google offers in terms of call screening and organizing is so much better than anyone else that it’s another reason I am loath to leave the Pixel line). All the “AI” stuff is weird and ethically problematic in one dimension or another, and all the other stuff just mostly works, so there’s your value proposition right there.

Now, we can argue what the substantive difference between “AI” and, say, “computational photography” actually is, because at the heart of both are a lot of the same processes and functionality, and that’s a reasonable argument to have. At this point, it really does seem like “AI” is “bullshit you don’t need or is done better in other ways, but we’ve just spent literally billions on this so we really need you to use it, even though it’s nowhere as good as what we were already doing,” and everything else is just unsexy functionality that makes what you do marginally easier or better. I’m sorry we live in a world where enshittification is being marketed as The Hot And Sexy Thing, but just because we’re in that world, doesn’t mean you have to accept it.

I do think it’s possible to oversell the doomsaying: The world will not end because we have democratized the ability to edit family snapshots outside of the realm of reality. But at the same time, please remind your parents and grandparents (and especially credulous friends) that the pictures they see on social media should no longer be assumed to have any relation to reality, especially when they are looped into a frantic/angry-sounding post about politics or social stuff, and that they shouldn’t forward that nonsense. If they continue to do so, mock them mercilessly for it, and then, probably, you should just block or mute them. After a certain point you have to assume people have agency, and they’re forwarding bullshit because they want to, not because they don’t know any better. Google can’t be (entirely) blamed for your racist uncle forwarding awful crap.

5. Back to the Pixel 9 Pro, is it a phone you should get? If you have a recent phone of reasonable quality, nah. Among Pixel users, if you’re a Pixel 7 or 8 user, you’re gonna be fine for another year. But if your phone is older than a couple of years and you’ve been thinking of doing an upgrade — and you’re looking in the Android sphere of things — then yes, it’s a really incredibly solid phone and I can recommend it (or, if you have bigger hands, the XL, or, if you don’t care about the ultrawide camera, the regular Pixel 9, which is a couple hundred dollars cheaper). I’ve not regretted having a Pixel phone, and kvetching about “AI” nonsensery aside, I’m at least initially very happy with this one as well. Google has its faults(!), but the Pixel phone line generally is not one of them.

— JS