lud_lite

In my experience they didn't. The managers did something. Witout them, senior coders become that layer of management.

They spend less time coding and more time managing.

So the company is paying for that organizing work all the same.

Probably overpaying for it, since a level up coder is now doing cross-level-down management and doing less architecture as a result.

Correct. This is a system design problem. You want this to be transactional and work at scale? That might be hard to achieve. Maybe if the data can be partioned then each node handles its own auditing in a table ad part of the transaction. There are many possibilities. Allowing inconsistently might be OK too depending on what is required.

What if you send the facts in the email. The facts that matter: request to book today as sick leave. Send that. Let the receiver run AI on it if they want it to sound like a letter to the King.

Even better. No email. Request sick through a portal. That portal does the needful (message boss, team in slack, etc.). No need to describe your flu "got a sore throat" then.

This exists (it is called Open AI, Anthropic etc.)

Thanks. I was asking more is this going to be a small business or do you see a 10M+ business?

Fixing things requires more finess than adding features. Because it needs to be fixed. Not fixed means you are out of business. Vibe coding is rolling a dice to see if something works. The question is whether these systems can stack that dice to converge to self healing within the time needed.

Maybe they will maybe they won't, or maybe it depends on the problem.

I think you need a expert human in the loop. Maybe fewer of them for a given weight of functionality.

That's is an interesting niche. I wonder if it is a lifestyle business or there is real potential for growth.

Stick the caps in series?

I am new to Amdahl's law, but wouldn't a rearchitecture make it less relevant. For example if instead of growing an LLM that has more to do in parallel, seperate it into agents (maybe a bit like areas of the brain?). Is Amdahls law just a limit for the classic LLM architecture?

If there I in AI then it would tell me where to put the hole and why.

Booleans also force the true/false framing.

E.g. a field called userCannotLoginWithoutOTP.

Then in code "if not userCannotLoginWithoutOTP or otpPresent then..."

Thus may seem easy until you have a few flags to combine and check.

An enum called LoginRequirements with values Password, PasswordAndOTP is one less negation and easier to read.

Another option is audit info could go to another table or datastore entirely.

If you never use it, that data can be dumped to s3 glacier periodically (e.g. after 90 days).

By losing the foreign key you gain flexibility in what you audit. Maybe audit the operation and not the 20 writes it causes.

Scrimba taught me React back in 2021. I recently checked if there was a Go course there, but sadly not. However with this announcement hopefully soon.

Why I like Scrimba? The lesson style that forces you to use a blank Page on each exercise means you develop memory for the language. Even more of a plus post LLMs IMO.

I'd be out of work entirely if I quit every job when I saw something disgusting. There are levels though.

Never ask me again is available here: https://www.debian.org/distrib/

My routine is to not use AI at all when writing.