Category: Uncategorized

Use (deep) neural networks to directly model total calories in an image.

For a given food:

1. Take photos from various angles and in various conditions
2. Weigh just the food—no containers, labels, or packaging
3. Blend up the entire food until it is a consistent puree
4. Use a bomb calorimeter or similar to find the number of calories in a gram of the puree
5. Multiply by total number of grams—this becomes the target for a convnet regression problem

Repeat for a ton of foods. Also repeat for non-foods since food-ness is part of the prediction. (Not JUST calories.)

Possibly constrain the convnet into one sub-network per macro component (protein, fat, alcohol, carbohydrate)

An SMTP server that requires lighting/bitcoin postage for every message it receives.

The service divides the postage with the recipient.

No spam filtering is done, leading to efficiencies.

The difficult part is not technical but social/economic: the network effect of non-postage SMTP is vast. By offering postage to any who receive the emails sent, participation could be incentivized. But adding a dependency on LN for the operation of the email system is a big/impossible ask. It would have to demonstrate substantial benefits over the status quo.

Track which counties you’ve visited

Highways driven

States

Countries

A robot mouse that learns to evade your cat.

Like a hierarchical notary system, transmitted via a gossip protocol

Every “knower” is vouched for by another “knower”, unless they want to start their own knowledge root.

Sort of like DNS, but instead of translating domain name -> IP, it vouches for general facts, or maybe translates hash -> person

Donations forwarded along the knowledge net, based on vouched-for facts?

Some system to demonstrate electronically that a public key is uniquely associated with a human being (a one to one relationship)

See also Concept: Maximum Basic Goodness 2

https://www.reddit.com/r/Unicode/comments/naafm9/handedness_modifiers

(Updates Concept: Maximum Basic Goodness; see also Concept: Proof of Person)

A blockchain-based system for voluntary wealth redistribution.

Components:

• Attestation of humanity: authorities that cryptographically attest that a given person is (within their system) a unique human being, not otherwise represented.
• Ability to publicly or anonymous donate to all people that meet a particular proof of humanness, e.g. every human attested by a state’s vital records division
• Ability to publicly (or anonymously?) subscribe to receive donations
• Ability to pledge certain amount and frequency of donations in the future

Need to allow distributed proof of humanness.

Various “humanity” authorities will make their own attestations.

Their attestations can be compared as well, i.e. they accept what another has rejected or vice versa

The first N times an assessment is made, the parties involved get a reward.

Or, the parties involved always get a reward, which declines according to a function, reflecting the diminishing returns involved.

(Could this reward be determined by a market?)

The reward mechanism implies that this is a currency.

The reward is the same for confirmations and rejections.

A high rate of rejection disincentivizes people to use the authority.

Confirmations can be annotated with useful information like genetic hashes and whatnot

(Closed or open vocabulary for annotations?)

Authorities can also attest to each other.

(Are authorities people?)

Like Web of Trust but with a financial incentive to get it right—hopefully. Or to do it at all.

Authorities (and people in general?) can publish contact information, allowing them to be investigated and their trust level increased.

https://www.reddit.com/r/cryptography/comments/bulgyu/cryptographic_systems_demonstrating_existence_of

I want there to simply be a “donate(amount)” function that equally distributes the amount to all humans on the system.

But that implies a default set of nodes that are valid.

You pick one or more “trust roots” (authorities) whose judgements you accept.

And you pick a minimum number of confirmations (default 2?) and maximum number of rejections (default 0)

Because received donations will be very small fractions, precision issues would easily arise. Donate 0.001 to a hundred million people and you start to see precision problems.

A routing protocol, applied generically to package delivery, such that anybody can participate in the delivery network

Cf. Lightning Network

Obvious issues, but fun to think about

Pull together https://todobackend.com/ in a command line app, a react native app, etc.

See also: Hecho (TODO link!)