Street View is such a missed opportunity. In 2007 it was visionary and essential to create the map data that allowed Google Maps to win. In 2026 it is a symbol of Google's stagnation. Essentially zero improvement in user experience for more than a decade, in a time of incredible advancements in computer vision.
By now we should all be flying around the planet in a seamless 3D reconstruction unifying street level and satellite views and allowing smooth free camera motion all the way from space to the front door of buildings and even inside. Many years ago I saw internal Google demos of dramatically improved Street View rendering, none of which ever made it to production. Google has consistently failed to recognize the value of the product and systematically underinvested in the user experience.
It makes sense they prototyped it. But putting it into production is $$$, way more expensive than current street view.
Current street view works well enough. How is a massively upgraded 3D version, that is bloated and slower to use on older devices, going to make Google more money?
It feels more like a separate product to license to architecture firms, city planners, video game studios, etc.
Knowing Google’s tendency to kill things they try and fail to revamp, I’ll take this stagnation as long as they keep updating it with new imagery. Street View is the greatest project in human historiography; there’s too much to lose to silly Google management.
We need an open version (as OpenStreetMap is to Google Maps).
Mapillary (https://www.mapillary.com/) has surprisingly good coverage in some places, but the experience is lacking, partly because most of the images (where I've looked) aren't 360 views.
No, you can't. Google Earth VR is indeed awesome (I am biased because I was involved in its creation), however there is no seamless integration between Street View and satellite view, and no free camera motion or even stereo rendering support for Street View in Google Earth VR. Google Earth VR was essentially abandoned and hasn't been updated at all since 2018, as can be seen in its Steam listing. This is due to the sad failure of the Daydream team.
Finally there is a glimmer of hope now that Android XR is happening. There is a new version of Google Maps for Android XR that does finally have a 3D reconstruction feature for Street View, but only for building interiors. Hopefully it won't be abandoned this time!
I’ve been pointing to Google Maps, drive as specific but not the complete set of fantastic innovation we saw around ~2007 for how great developers used to be.
I think the drift is specifically tied to the introduction of leetcode in the interview process. Which may sound like a wild connection at first but I’ve now lived through being blocked and seeing how creative devs can’t get through leetcode gatekeepers who are microfussing and blanket critiquing devs as bad when they don’t have leetcode answers pre memorized in a mental hasmap to be able to regurgitate from memory which allows the extra mental capacity to free up in order to hold a performative class lecture about it at the same time.
You can spend your time memorizing the test taking skills to be good at tests. Maybe memorize the answers too. Or you can be coming up with grand ideas like maps and street view and thinking about how all these things in the world come together to be able to do that.
Not many are good at both and the entire stack of people doing interviews is currently blocked at fixing this. Nobody wants to have wasted their time memorizing leetcode to just not gatekeep people who didn’t put in “the same effort,” and no hiring team wants to gamble on somebody who fails the leetcode test processes and turns out to be the occasional bad hire with the only paperwork saying they didn’t pass the industry standard test and shouldn’t have gotten hired in the first place.
So we’re now blocked with only slop workers getting hired who don’t feel the same comfort to take big risks and we get slop like Microsoft notepad plus copilot 365 as a result.
Was leetcode-style interviewing not a thing before that? Cracking the coding interview was published in 2008 so I would assume it was already pretty quite established by then.
My first job, in California, was just transitioning to leetcode-style interviews in 2007 following the industry trend. So it was probably spreading around that time.
While I agree something like this sounds really neat, I am curious what the value proposition is? Pointedly, is it any higher than doing the same thing in a video game in a fantasy world?
The difference is that it's useful for navigating the real world. You could have way better directions displays that show directions in context instead of just schematically. It would make the petabytes of imagery that has already been collected much more accessible and therefore useful, instead of being relegated to a special clunky Street View mode that is rarely visited. It would enable exploring real spaces in a way that provides much better spatial context, to build a spatial memory that helps your navigation when you get to the real place. And yes, it would be fun. At one time, Google was into that sort of thing.
I could see this as an argument for a heads up display. So, good for projecting directions onto a windshield or for having the glasses thing. But this? I don't see how a VR world helps anyone navigate the real world. That is, you seem to be saying the VR data is needed for AR usage. And I just don't see how those are helping each other too much.
I'm fully bought off on the "it would be fun" aspect. I don't see a value proposition for it, though.
A heads up display doesn't need a 3D rendering of the environment around it because the environment is already visible through the screen. The 3D rendering is so you can see what to expect before you get there. If you don't understand why that could possibly be useful then I don't know what to tell you; you'll have to take it for granted that some people's brains work differently than yours and can benefit from seeing places they are about to visit in 3D before they get there.
Apologies, I meant my point to be that navigating a place is more helped with AR techniques than it is VR ones. Which, as you say here, is less helped by 3D rendering than it is other things. Indeed, I meant that to be my point.
Do I think it could be useful if you rehearse navigating a place before getting there? Yeah. Ish. I can see obvious military style value adds for that. Average person, though? I still have a hard time seeing the value.
Reminds me when places were offering video tours of places. Is a neat idea. But ridiculously low in actual value.
Reading a map isn't that hard. It just sounds like an elaborate way to illustrate navigation with crayons. A cool product demo, but not very useful in practice.
What you describe seems to have been implemented in Google Earth. It seems like an intentional product choice to do it in Google Earth and not Google Maps. Most people use Google Maps to get directions and reviews of places, and very few people I know even use the street view feature.
Google Earth and Google Maps 3D satellite view on desktop web have essentially the same rendering of Street View, which hasn't materially changed in 10 years or more and does not have the features I described. Google Maps on mobile never integrated the 3D satellite view at all, which represents another regrettable lack of investment on Google's part.
People rarely use the Street View feature because it's difficult to access, and difficult to understand spatially. Free camera motion is impossible and the transitions are jerky and stilted. As a result it's relegated to special places in the UI that are rarely visited. If it was seamlessly integrated into satellite view and directions then it would see much more usage.
There's a parallax effect in Street View on Apple Maps that separates out the layers of every image. Things like lampposts or telephone poles all rotate slightly differently to whatever is behind them.
And it's such a subtle effect that I still break my brain trying to determine whether or not I've made it up.
Imagine expending that much development time and effort for something you're not even sure is there. And somehow I still find it enviably cool.
I think it's the same tech they use to make the "3d" background photos on the iPhone wallpaper, which is probably also the same tech used for inferring depth when converting a normal photo to a spatial photo for viewing on an AVP.
I just wish Apple would add more streets. London is the closest place with Apple “street view” to me and there are literally a couple of cities (!!!) between me and London. So I don’t hold any hope that Apple will ever get round to coming to my small village if there are entire cities they’ve left out.
Sounds awesome, is there a video of it I could watch somewhere?
Edit: for those who didn't know, like me, apple's maps are available at https://maps.apple.com. You can see this yourself. The effect is unvelievably smooth compared to what Google maps have
The workstation paragraph seems like a humble brag. Most of us yearn for a set-up like that! Especially with the price of components going up thanks to AI and corporations buying all the hardware to support it.
The author is known for deep dives on data sets like that (I'm following him on Linkedin for that), so makes sense they always mention their setup even if it doesn't apply to his specific data set.
It is a humble brag. I saw the specs and thought the author would discuss different approaches of finessing the data and a benchmark. There isn’t one. So it’s indeed a humblebrag.
Tangential comment but I still don't understand how we have technology to identify a car license plate from space but we have pixelated images from Antarctica on Google Maps / Google Earth. Why not publish that and make it accessible? Is it true that Antarctica is not easy to scan due to ice and snow?
Not sure this is the reason but: it is generally not easy to get a satellite over the poles. You launch close to the equator in the direction of Earth's spin to take advantage of the (very substantial!) speed you already have due to the rotation of the planet. Getting from an equatorial orbit to a polar one requires a huge amount of fuel / energy. You can't just sort of "drive it over".
Google Maps' high-resolution "satellite" imagery is actually captured from planes.
Antarctica is huge (1.5x the area of the US), it would be a dangerous logistical nightmare to fly the sorts of patterns you need to capture aerial imagery there, and it's almost entirely covered in non-descript ice -- what would be the value of having high-resolution imagery there?
I wonder if we just don't have many of these types of satellites in a polar orbit, since we don't have as big a need for that type of imagery for the poles?
The high-res imagery on Google Earth mostly comes from aerial surveys, not satellite. If it's not economically worth flying a plane back and forth to survey -- and that's certainly true in Antarctica -- that's when you get the fuzzy civilian-grade satellite imagery or some cheap/public-domain out-of-date aerial photography.
Costa Rica and Paraguay coverage was added recently (within the last year iirc). The author notes Paraguay as an example of a country that was not yet in the dataset they sourced from.
El Salvador does have a decent amount of coverage on street view, but this was done by El Salvador Maps (if you pan the camera down, you'll see this name on the cars used to capture the coverage). The dataset is curated by a member of the Geoguessr community, in which "unofficial" coverage like this is disregarded, which is why you won't see it included.
That whole part of Ontario is basically farms with long straight concession roads so I imagine you could cover a lot of area quite quickly just driving in a straight line for a couple hours, turning driving 2 km then turning and driving straight for a couple hours on repeat.
"The darker colours are points that were updated closer to 2007 and the brighter colours closer to December of last year." It's possible that this area was just more recently updated and is not necessarily more densely covered compared to other areas.
Streetview is such an incredible product - one of the few digital products that still manages to bring me joy every day. it'll be a shame when it's inevitably enshitified.
I can see the value of it, certainly, but it's also probably Google's creepiest product. The street where I live, you can see inside peoples kitchens and living rooms on Street View. I had to ask Google twice to block my house, because they fucked it up the first time.
Street View is such a missed opportunity. In 2007 it was visionary and essential to create the map data that allowed Google Maps to win. In 2026 it is a symbol of Google's stagnation. Essentially zero improvement in user experience for more than a decade, in a time of incredible advancements in computer vision.
By now we should all be flying around the planet in a seamless 3D reconstruction unifying street level and satellite views and allowing smooth free camera motion all the way from space to the front door of buildings and even inside. Many years ago I saw internal Google demos of dramatically improved Street View rendering, none of which ever made it to production. Google has consistently failed to recognize the value of the product and systematically underinvested in the user experience.
We could be... but why? What's the product?
It makes sense they prototyped it. But putting it into production is $$$, way more expensive than current street view.
Current street view works well enough. How is a massively upgraded 3D version, that is bloated and slower to use on older devices, going to make Google more money?
It feels more like a separate product to license to architecture firms, city planners, video game studios, etc.
Wait until you find out how much money they are burning so you can generate an image of a fake banana.
Knowing Google’s tendency to kill things they try and fail to revamp, I’ll take this stagnation as long as they keep updating it with new imagery. Street View is the greatest project in human historiography; there’s too much to lose to silly Google management.
We need an open version (as OpenStreetMap is to Google Maps).
Mapillary (https://www.mapillary.com/) has surprisingly good coverage in some places, but the experience is lacking, partly because most of the images (where I've looked) aren't 360 views.
Is this actually open source? The few datasets I found on there are under non-commercial licenses.
You can do that with Google Earth VR. It’s actually really cool in VR. You feel like Godzilla.
Unfortunately it’s only a small subsets of major cities and the implementation feels so half-baked it could have been an AWS service.
But it’s still a cool tech demo nonetheless.
No, you can't. Google Earth VR is indeed awesome (I am biased because I was involved in its creation), however there is no seamless integration between Street View and satellite view, and no free camera motion or even stereo rendering support for Street View in Google Earth VR. Google Earth VR was essentially abandoned and hasn't been updated at all since 2018, as can be seen in its Steam listing. This is due to the sad failure of the Daydream team.
Finally there is a glimmer of hope now that Android XR is happening. There is a new version of Google Maps for Android XR that does finally have a 3D reconstruction feature for Street View, but only for building interiors. Hopefully it won't be abandoned this time!
I’ve been pointing to Google Maps, drive as specific but not the complete set of fantastic innovation we saw around ~2007 for how great developers used to be.
I think the drift is specifically tied to the introduction of leetcode in the interview process. Which may sound like a wild connection at first but I’ve now lived through being blocked and seeing how creative devs can’t get through leetcode gatekeepers who are microfussing and blanket critiquing devs as bad when they don’t have leetcode answers pre memorized in a mental hasmap to be able to regurgitate from memory which allows the extra mental capacity to free up in order to hold a performative class lecture about it at the same time.
You can spend your time memorizing the test taking skills to be good at tests. Maybe memorize the answers too. Or you can be coming up with grand ideas like maps and street view and thinking about how all these things in the world come together to be able to do that.
Not many are good at both and the entire stack of people doing interviews is currently blocked at fixing this. Nobody wants to have wasted their time memorizing leetcode to just not gatekeep people who didn’t put in “the same effort,” and no hiring team wants to gamble on somebody who fails the leetcode test processes and turns out to be the occasional bad hire with the only paperwork saying they didn’t pass the industry standard test and shouldn’t have gotten hired in the first place.
So we’re now blocked with only slop workers getting hired who don’t feel the same comfort to take big risks and we get slop like Microsoft notepad plus copilot 365 as a result.
Was leetcode-style interviewing not a thing before that? Cracking the coding interview was published in 2008 so I would assume it was already pretty quite established by then.
My first job, in California, was just transitioning to leetcode-style interviews in 2007 following the industry trend. So it was probably spreading around that time.
While I agree something like this sounds really neat, I am curious what the value proposition is? Pointedly, is it any higher than doing the same thing in a video game in a fantasy world?
The difference is that it's useful for navigating the real world. You could have way better directions displays that show directions in context instead of just schematically. It would make the petabytes of imagery that has already been collected much more accessible and therefore useful, instead of being relegated to a special clunky Street View mode that is rarely visited. It would enable exploring real spaces in a way that provides much better spatial context, to build a spatial memory that helps your navigation when you get to the real place. And yes, it would be fun. At one time, Google was into that sort of thing.
I could see this as an argument for a heads up display. So, good for projecting directions onto a windshield or for having the glasses thing. But this? I don't see how a VR world helps anyone navigate the real world. That is, you seem to be saying the VR data is needed for AR usage. And I just don't see how those are helping each other too much.
I'm fully bought off on the "it would be fun" aspect. I don't see a value proposition for it, though.
A heads up display doesn't need a 3D rendering of the environment around it because the environment is already visible through the screen. The 3D rendering is so you can see what to expect before you get there. If you don't understand why that could possibly be useful then I don't know what to tell you; you'll have to take it for granted that some people's brains work differently than yours and can benefit from seeing places they are about to visit in 3D before they get there.
Apologies, I meant my point to be that navigating a place is more helped with AR techniques than it is VR ones. Which, as you say here, is less helped by 3D rendering than it is other things. Indeed, I meant that to be my point.
Do I think it could be useful if you rehearse navigating a place before getting there? Yeah. Ish. I can see obvious military style value adds for that. Average person, though? I still have a hard time seeing the value.
Reminds me when places were offering video tours of places. Is a neat idea. But ridiculously low in actual value.
Reading a map isn't that hard. It just sounds like an elaborate way to illustrate navigation with crayons. A cool product demo, but not very useful in practice.
That's monopoly behavior, baby! Break 'em up.
Website named Hackernews: “I’d like the government to forcibly dismantle the organizations that produce the vast majority of technical advancements”
No it's not.
What you describe seems to have been implemented in Google Earth. It seems like an intentional product choice to do it in Google Earth and not Google Maps. Most people use Google Maps to get directions and reviews of places, and very few people I know even use the street view feature.
Google Earth and Google Maps 3D satellite view on desktop web have essentially the same rendering of Street View, which hasn't materially changed in 10 years or more and does not have the features I described. Google Maps on mobile never integrated the 3D satellite view at all, which represents another regrettable lack of investment on Google's part.
People rarely use the Street View feature because it's difficult to access, and difficult to understand spatially. Free camera motion is impossible and the transitions are jerky and stilted. As a result it's relegated to special places in the UI that are rarely visited. If it was seamlessly integrated into satellite view and directions then it would see much more usage.
Oh I thought the desktop 3D satellite view was powered by combining the regular satellite view and street view. Looks like that’s not the case.
Apologies as this is fairly tangential:
There's a parallax effect in Street View on Apple Maps that separates out the layers of every image. Things like lampposts or telephone poles all rotate slightly differently to whatever is behind them.
And it's such a subtle effect that I still break my brain trying to determine whether or not I've made it up.
Imagine expending that much development time and effort for something you're not even sure is there. And somehow I still find it enviably cool.
Oh wow I hadn’t noticed that. It’s especially noticeable when “walking” down the street between points.
I’m so glad to finally have that feature in my area. It was the one thing I missed from Google Maps which I otherwise avoid.
I think it's the same tech they use to make the "3d" background photos on the iPhone wallpaper, which is probably also the same tech used for inferring depth when converting a normal photo to a spatial photo for viewing on an AVP.
I just wish Apple would add more streets. London is the closest place with Apple “street view” to me and there are literally a couple of cities (!!!) between me and London. So I don’t hold any hope that Apple will ever get round to coming to my small village if there are entire cities they’ve left out.
I love this and share the experience! It is a very cool effect, specially when moving through the street.
Sounds awesome, is there a video of it I could watch somewhere?
Edit: for those who didn't know, like me, apple's maps are available at https://maps.apple.com. You can see this yourself. The effect is unvelievably smooth compared to what Google maps have
Oh cool, I didn't know they added their Street View equivalent to the web version. The animated transitions are much better than Google's.
It does look smooth but for me the killer feature on google maps is still the ability to go back in time.
They seem to be significantly slower though
Does it appear only if you visit the website on mobile? I don't see any street view option when I visit that site on a PC.
Oh that’s right, I always forget they have it on the web.
Drag two fingers on the map and you'll bring the camera down to see the 3D effect more clearly.
For my life, I don't understand the gestures for the 3D map. It would seem that I can ONLY manipulate that view by accident.
The workstation paragraph seems like a humble brag. Most of us yearn for a set-up like that! Especially with the price of components going up thanks to AI and corporations buying all the hardware to support it.
The author is known for deep dives on data sets like that (I'm following him on Linkedin for that), so makes sense they always mention their setup even if it doesn't apply to his specific data set.
Yea, I agree. The dataset is < 100MB... so duckdb can very easily handle this on an old macbook air.
https://duckdb.org/2025/05/19/the-lost-decade-of-small-data
It is a humble brag. I saw the specs and thought the author would discuss different approaches of finessing the data and a benchmark. There isn’t one. So it’s indeed a humblebrag.
Tangential comment but I still don't understand how we have technology to identify a car license plate from space but we have pixelated images from Antarctica on Google Maps / Google Earth. Why not publish that and make it accessible? Is it true that Antarctica is not easy to scan due to ice and snow?
Not sure this is the reason but: it is generally not easy to get a satellite over the poles. You launch close to the equator in the direction of Earth's spin to take advantage of the (very substantial!) speed you already have due to the rotation of the planet. Getting from an equatorial orbit to a polar one requires a huge amount of fuel / energy. You can't just sort of "drive it over".
Source: played a bunch of Kerbal space program
Google Maps' high-resolution "satellite" imagery is actually captured from planes.
Antarctica is huge (1.5x the area of the US), it would be a dangerous logistical nightmare to fly the sorts of patterns you need to capture aerial imagery there, and it's almost entirely covered in non-descript ice -- what would be the value of having high-resolution imagery there?
I wonder if we just don't have many of these types of satellites in a polar orbit, since we don't have as big a need for that type of imagery for the poles?
Pretty sure the license plate/street view level data is from the cars that drive around with a lidar thing strapped to the top, not satellite data
The high-res imagery on Google Earth mostly comes from aerial surveys, not satellite. If it's not economically worth flying a plane back and forth to survey -- and that's certainly true in Antarctica -- that's when you get the fuzzy civilian-grade satellite imagery or some cheap/public-domain out-of-date aerial photography.
El Salvador looks black but has pretty good coverage throughout the country.
Costa Rica seems also to have more coverage than I see here.
Paraguay too.
Costa Rica and Paraguay coverage was added recently (within the last year iirc). The author notes Paraguay as an example of a country that was not yet in the dataset they sourced from.
El Salvador does have a decent amount of coverage on street view, but this was done by El Salvador Maps (if you pan the camera down, you'll see this name on the cars used to capture the coverage). The dataset is curated by a member of the Geoguessr community, in which "unofficial" coverage like this is disregarded, which is why you won't see it included.
Why is there so much dense coverage in southern Ontario than anywhere else in North America?
An unusually dense road network?
edit: This page has some data: https://www.researchgate.net/figure/Landscape-metrics-for-ro...
Southern Ontario has 4x the road density of the province average, so might be a contributing factor?
That whole part of Ontario is basically farms with long straight concession roads so I imagine you could cover a lot of area quite quickly just driving in a straight line for a couple hours, turning driving 2 km then turning and driving straight for a couple hours on repeat.
"The darker colours are points that were updated closer to 2007 and the brighter colours closer to December of last year." It's possible that this area was just more recently updated and is not necessarily more densely covered compared to other areas.
Is that base map style inspired by the board game Pandemic, the computer game DEFCON, or a third thing?
Edit: Apparently it is "Nova Map" base tile set from ArcGIS.
Streetview is such an incredible product - one of the few digital products that still manages to bring me joy every day. it'll be a shame when it's inevitably enshitified.
I can see the value of it, certainly, but it's also probably Google's creepiest product. The street where I live, you can see inside peoples kitchens and living rooms on Street View. I had to ask Google twice to block my house, because they fucked it up the first time.
> it'll be a shame when it's inevitably enshitified
Depending on where you live, that happened about 10 years ago.