> Cataloguing your fridge requires taking pictures of everything you add and remove which seems... tedious. Just remember what you have?

I agree that removing items and taking pictures takes more effort than it saves, but I would use a simpler solution if one existed because it turns out I cannot remember what we have. When my partner goes to the store I get periodic text messages from them asking how much X we have and to check I look in the fridge or pantry in the kitchen and then go downstairs to the fridge or pantry in the basement.

> Can you not prepare for the next day by opening your calendar?

In the morning I typically check my work calendar, my personal calendar, the shared family calendar, and the kids' various school calendars. It would be convenient to have these aggregated. (Copying events or sending new events to all of the calendars works well until I forget and one slips through the cracks...)

> If you have reminders for everything (responding to texts, buying gloves, whatever else is not important to you), don't you just push the problem of notification overload to reminder overload?

Yes, this is the problem I have. This doesn't look like a suitable solution for me, but I understand the need.

> In the morning I typically check my work calendar, my personal calendar, the shared family calendar, and the kids' various school calendars. It would be convenient to have these aggregated. (Copying events or sending new events to all of the calendars works well until I forget and one slips through the cracks...)

But... calendar apps already let you aggregate your calendars into a single view. Even if you have them on separate accounts (or some other impediment), you can easily share a read-only version of, say, your work calendar with your personal account so that you can have them combined in the morning.

> you can easily share a read-only version of, say, your work calendar with your personal account so that you can have them combined in the morning.

If only it was that easy! I'm not allowed to share content to or from my work calendar for security reasons. The school and camp calendars are a mix of PDFs and hand-written websites -- a neighbor wrote a scraper to extract the information from a few of them into a caldav at one point but it ended up being even flakier than copying the relevant bits by hand. There's no technical barrier to consolidating my personal calendar with the various family / neighborhood calendars but in practice I have to hide most of the other calendars because the volume of irrelevant events is just too large, so I end up just copying over the relevant events to a personal calendar.

I think this problem is one that AI could actually help with- simply snap a photo of my school calendar and ask the ai to add the important items to my personal calendar.

But I don't need the AI to do this everyday, just when i get a new calendar.

It honestly tempting to point a camera at my workstation so AI can "watch over my shoulder" while I'm working on systems that are pointlessly excessively locked down.

I don't do, but it is tempting, and I bet people will do it.

Too much security makes people seek insecure workarounds...

Or to quote Star Wars, “The more you tighten your grip, Tarkin, the more star systems will slip through your fingers”.

Many years ago I watched someone marched out of the room in handcuffs by military police for plugging a USB thumb drive in the wrong computer.

My current situation isn't anywhere near that strict, and I agree that many security postures are dumb and overbearing, like unnecessarily frequent password rotation. But honestly, preventing employees from sharing company documents with random third parties doesn't seem all that unreasonable.

I agree, but a lot of companies risk exactly that by creating policies that people are likely to have reasons to want to bypass.

E.g. Calendar sharing. It's a paintpoint if you often have irregular working hours and have to match up a personal and work calendar. At least allow sharing busy/not busy... By not doing so, you create an environment where people are tempted to find workarounds that might be much worse.

Part of your security posture needs to be to consider how to prevent friction in areas where reducing it removes incentives for non-compliance.

So Clawdbot has access to your work calendar that’s supposed to be secure ?

No, that's why I said "This doesn't look like a suitable solution for me, but I understand the need."

We have forgotten the simple, reliable solutions of the past - a grocery list on the fridge, a weekly planner, a weekly plan itself rather than constant coordination. Cell phones and easy communication led us here.

I'm curious what makes you think the solutions of the past have been forgotten or that they were somehow more reliable? (They're certainly simpler, I'll give you that!)

I have printouts of school/camp calendars taped to the wall, a weekly planner on the kitchen whiteboard, paper grocery lists on the fridge, and a pocket notebook for capturing random tasks. I used to believe that some lifehack, process, methodology, app, or modern jeejah would finally solve my organization problems. But as I got older I made peace with the fact that they're all limited by the same weak link -- me.

Ignoring the fact that OP does not know about existing solutions like Grocy where people do find value in the currently tedious setup of adding products and tracking their kitchens inventory, and just zeroing in on your first point. The paper grocery list is terrible

If you cook at all a solution like Mealie becomes your cookbook. Its trivial to create grocery list for when you take the time to plan out your meals for the day, week, or month. If you are not shopping by yourself, everyone on the app can just pick up things in the grocery store independently. Its an actual time saver

Mealie exposes an API so you could theoretically expose it to another solution like Home Assistant and have your grocery list sync with your errands list. Suddenly you have the ability so that anyone using Home Assistant could get an alert when they are nearby the grocery store or Costco to pick up things on this combined list. Maybe your partner is walking by a store you've created a zone for, with items on your master list, gets an alert, and they can mark off some things that they picked up and it syncs back to where the items were originally added. Your inventory is then updated based on marked off items.

Now imagine if you did not have to come up with the bespoke master list for all the stores you go to and it can determine when to send that alert. You can also just snap a picture of your receipt or shopping cart and it is all figured out for you.

But you could just use piece of paper with the magnet on the fridge.

Theres a lot of manual process that can be eliminated for things people already find convenient enough to do manually. Local models can easily handle much of this already.

> In the morning I typically check my work calendar, my personal calendar, the shared family calendar, and the kids' various school calendars. It would be convenient to have these aggregated. (Copying events or sending new events to all of the calendars works well until I forget and one slips through the cracks...)

Why in the world would you use a non-deterministic system for something so banal but important?

LLMs regularly let things slip through the cracks in ways no human would ever do so.

> Why in the world would you use a non-deterministic system for something so banal but important?

I wouldn't. As mentioned above, this (using an LLM) doesn't look like a suitable solution for me, just pointing out that I understand the need.

> When my partner goes to the store I get periodic text messages from them asking how much X we have and to check I look in the fridge or pantry in the kitchen and then go downstairs to the fridge or pantry in the basement.

We used to have a similar problem until we made a policy that if you use something up you add it to our shared shopping list, usually with a voice command to Siri. Whenever someone is at the store we just check the list, making sure we mark off things that are purchased.

Officially we have a similar policy except that it's a paper list next to the pantry. But with a half-dozen people in our household the likelihood that everyone has been 100% reliable in adding finished items to the list and there are no omissions is low, hence the text messages.

> but I would use a simpler solution if one existed

I often thought about a magnetic barcode scanner that is attached to the fridge and connected to some form of inventory app, but it would be useless at fresh produce without a barcode.

Fridge cataloging is actually a great use case for image recognition, the problem is fridges no accommodations to power accessories inside them.

I have a couple of temperature sensors to alert Home Assistant if the fridge gets too warm. It would be easy and cheap to add some ESP32-camera modules to track contents...but there's no way to power them nicely (I simply don't know where I could pull USB power through).

Samsung makes an "AI Vision" fridge I looked at briefly, but it didn't come close to making sense for us given the unreliability of the vision system, the cost of replacing a couple fridges, and the comparative simplicity of a paper list.

I have one. It’s the stupidest thing ever. It tries to detect food items going in and out, requires confirmation on screen, and maybe categorises 5% of things automatically.

Very very very flat cables don't mess with the gaskets on the door too much.

You can only track what containers are in the fridge, not how much is left or if it’s expired. “Automated” pantry or fridge tracking is just not possible and requires way more effort than just writing “mustard” on the shopping list when you notice you’re low.

If you had a scale with an image recognition camera and you put everything on the scale before and after removing it from the fridge, it would probably work pretty well? I've been pondering setting something like that up, it would also be really helpful for keeping track of how much and of what I'm actually putting into the food I made, if I weigh everything before and after, I can just collect the amounts after the fact and don't have to worry as much about measuring if I want to make the same dish again.

That would be extremely cool, I would love to follow you on youtube or something if you decide to pursue this.

Again, you still have to put in way more work. You have to somehow know the weight of the container, otherwise it will never register as empty. Or you have to know the volume and the density of its contents (or worse, think about a jar of olives or pickles, how would a weight tell you it’s empty with the brine in there?). You still don’t know the expiry date. There’s no chance of automatically tracking this stuff.

The weight and expiry of the contents is printed on the package. Brine is a problem but olives and pickles it's much more tractable to estimate from a picture. These are all essentially solved image recognition problems. It doesn't need to be perfect to be useful either. The expiration date is the trickiest one, but mostly because you would need cameras on all sides including top and bottom, so you might end up having to hold it up for a moment to make sure the bottom is clearly photographed.

And yes, it's a bit more work but it gives high-fidelity data, with the right software you could calculate your actual nutrient intake with very high fidelity, which would actually be worth an extra 15-30 minutes a day of effort.