bootywizard

Karpathy notes that the model, or specifically it's DeepSearch feature "doesn't like to reference X by default" which seems counter to this.

Hopefully that means it is a joke...

I've played Mafia several times and enjoy it a lot. However, I have also witnessed friendships completely destroyed in the process. Some people are capable as seeing it as just a game, discard all prior trust or expectations with others during, and then at the end, reset completely back to how it was before, perhaps having learnt something about people in the process.

For those who cannot do this, they will experience true pain, broken trust, and leave with friendships fundamentally changed. If this sounds like you, do not play this game!

Hi Francois, congrats on leaving Google!

ARC and On the Measure of Intelligence have both had a phenomenal impact on my thinking and understanding of the overall field.

Do you think that working on ARC is one of the most high leverage ways an individual can hope to have impact on the broad scientific goal of AGI?

How an artificial neuron learns*

Whether you're trying to do this for a data warehousing use case, or in an environment where you are creating production database queries will change the approach you need to take here.

We have been working on SQLX for the data warehousing use case, allowing you to embed JavaScript into your SQL queries and making things like code re-use much easier while fitting in with your existing SQL dialect, might be of interest - https://docs.dataform.co/guides/sqlx. Some of the concepts in there are specific to our framework, some are not.

For data warehousing, create more views! They are IMO the right level of abstraction for encapsulating common business logic, data definitions, joins etc, making composing downstream queries much easier. Managing lots of views like this requires some investment in a data modelling tooling tool however (Dataform, DBT etc).

For generating queries used against production databases for building user applications, none of this applies.

Author here - it does but it doesn't, which is why it doesn't always work so well. It never actually keeps track of the shapes in a seperate data model - the text itself is the only data model. When you resize, it retraces out the lines that need editing. I did this as the expected use case is for diagrams in code, where diagrams may be copy pasted in and out of the tool and I'd lose any underlying models that I was working with. For fun: https://github.com/lewish/asciiflow2/blob/master/js-lib/draw...