Very interesting. But I wonder how much Google (and other) Maps can actually shape the scene. For tourist hotspots with a lot of visitors, it IS clearly the driving force. But for locals, I don’t think it has an overwhelming effect. Locals know their restaurants and they visit them based on their own rating. They could explore total strange and new ones, but then they will form their own rating and memory immediately and will not get fooled/guided by algorithm (the next time)

Yeah, can’t comment about London, as I’ve only been a tourist there, but assuming it works like in Tokyo. In a big city, with basically unlimited amount of dining options, a lot of people will try different places. In the past year, I don’t think I’ve repeated a single dinner spot more than 3 time, and I basically eat out every day. This is always a discovery problem, and word of mouth/google maps/tabelog/etc. is a major sales driver here.

Now, if I think about the time I lived in Vancouver, it was the opposite. You don’t have that many options, after a while you basically make a list of your favourites and rotate.

Long-time Tokyo/Yokohama resident here. I’m basically the same: Especially if I’m by myself and near a train station or retail area, I just walk around to see what’s available and choose someplace to eat. Only if I am planning a meal with others do I look for options online, and then, in addition to Google and Apple maps, I also use sites such as tabelog.com and restaurant.ikyu.com.

I haven’t been outside Japan for nearly a decade so I can’t compare it with other countries, but my impression is that Japan has more small restaurants than some other places. It’s not unusual to go into a ramen, curry, gyoza, soba, or other eating place with fewer than a dozen seats and staffed by just one or two people.

The existence of such small places increases the eating-out options. I don’t know why such small food businesses are viable here but not elsewhere; perhaps regulatory frameworks (accessibility, fire, health, tax, labor, etc.) play a role.

Totally. We’re definitely lucky over here. From my talks with people in restaurant industry in NA, it’s just extremely expensive to start a business, on top of the regulatory restrictions that you’ve mentioned. And obviously the holy grail of money making - liquor. I can get beer in almost every random ramen shop near me. It takes months/years of approval to open a place with a liquor license in Vancouver, Canada. Margins on alcohol are huge, that gives breathing room to little margins places make from food.

Unless, as a local looking for new spots to try, your first step is going to Google Map and searching "restaurants". I'm certainly guilty of this sometimes.

I did exactly this < 10 minutes ago. For my local area.

I disagree, i’m always using Google to find new restaurants and places to go to in my own (fairly large) city.

The writer is in London where even locals often eat outside their immediate neighborhood.

Depends if you live in a big city with a lot of restaurant turnover or not.

This is actually a big frustration for me how I can search food and get totally different results over the same area in the frame. I seem to remember in the old days of google maps you'd see, you know, everything in the area. Like pins on pins on overlapping pins. And you'd click through them or zoom in as appropriate. You found everything. It all worked.

Then someone had the brilliant idea that this was all too busy, and you should have pins omitted until you have sometimes zoomed so far in you are filling your map viewer frame with the doorstep of that business...

I wouldn't be surprised to learn businesses get charged to appear first. Seems like it tends to be things like fast food or national chains over new locally owned restaurants that pop up more often on google maps.

I'm not sure the overlapping pins idea would work for e.g. a 5 floor building with no multilevel maps and 6 businesses to a floor. Which is a common thing in some of the places Google maps.

Works for me. If I search "restaurants", and I see a building full of pins. I can now go to that building and look at all the restaurants.

You don't want to show every business as a default view, of course.

I still want to read what the poster understood from the output of the AI, though. I don’t need reciting an answer from an AI because I (and everybody else) can do it, too. On Firefox and other browsers, it’s now integrated so asking an AI is no more than 1 click away. Actually, not even away, Grok can even answer right in the context on X. So merely an answer from AI had no value today, whatsoever.

I think a helpful conclusion is that while the firing pattern in organoids doesn’t preclude a wetware of complex programmed instructions, it could be just the emergent properties of the underlying physics and electrochemical properties of the neurons; analogous to the phenomenon of synchronism when placing pendulums in a common place.

Delete cookies and site data on Firefox works.

I don’t know. I cannot even answer most of these questions straightforward with a or b!

These problems are well known for a long time, especially if one simply asks LLM for a changing fact, such as who is the current pope. But there is also a simple technique that reduces these issues almost to zero: thinking and explicit request of grounding. For example, asking any LLM: who is the current pope could give a wrong answer due to the fact that Pope Francis died in April 2025 then the cut-off date of these models may be before that date. A simple question triggers simple associations, and so the answer could be wrong. But if turn on the thinking mode and instruct for grounding, the LLM will answer correctly.

For the above example, asks instead: "Who is the current pope? Ground your answer on trustworthy external sources only" with thinking mode on or explicitly "think harder for better answer", all popular AI (ChatGPT 5+, Gemini 2.5 Flash, Claude 4+, Grok 4+) will answer correctly, albeit with sometimes long thinking time (28 s by ChatGPT 5 for example).

Without explicit instructions, the accuracy of the result depends heavily on the cut-off date and default settings of each model. Grok 4, for example, in auto-mode will do a search then answer correctly, but Grok 3 will not.

>> Unfortunately, Google seems to have the ability to arbitrarily flag any domain and make it immediately unaccessible to users. I'm not sure what, if anything, can be done when this happens, except constantly request another review from the all mighty Google.

Perhaps a complaint to the ETC for abusing the monopoly and lack of due process to harm legitimate business? Or DG COMP (in the EU).

Gather evidence of harm and seek alliances with other open-source projects could build a momentum.

If what you said is true, I wonder whether (fine-tuned AI) will replace mediocre scientists and specialists very soon.

I think changing the headline like that is not encouraged. If one find the fact that ChatGPT loses half of its paid subscribers, one should perhaps write an own article with data and other facts and not changed the headline of a linked article. Or did the site change it themselves?