One of the easiest ways to network with high-level individuals, is to just pay them.

In ML, this probably means taking their courses and going to their office hours/live calls (which most seem to have).

And this is way cheaper!

The identity of being a recent grad is worse than actually being a recent grad.

Some of these job interviews could literally just be an email.

MLOps tooling is more important than ever.

As we head into the new year,

I just want to remind you,

R is for the unemployed.

Why do people do this?

Machine learning jobs that have the "AI Engineer" title should be called LLM Engineer instead.

Leave this hyped up non-sensical title to the people building GPT wrappers.

The AI Engineer is the worst role in tech.

Because no one knows what it is or skills it entails.

It could be a software dev position with LLM APIs in the mix, or it could be more ML focused where you'd be fine-tuning.

Or something in the middle.

And don't forget those clueless recruiters too.

Being a Physics major is a skill issue.

Worst thing you can do in a competitive market

or really any tech market

is be a recent grad

so don't

There are definitely not enough machine learning engineers on this platform.

I hate anacondas

𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀𝗲𝘀 𝗱𝗼𝗻’𝘁 𝘄𝗮𝗻𝘁 𝗔𝗜—𝘁𝗵𝗲𝘆 𝘄𝗮𝗻𝘁 𝗿𝗲𝘀𝘂𝗹𝘁𝘀.

It's tempting to just throw in "AI" everywhere in marketing material, and let the hype sell for you.

But people only really care about how they can:

- Make money
- Save money
- Save time

Sell results, not AI.

Chatbots and Fine-Tuning are two of the most popular LLM applications out there.

Yet I think in a practical sense, your first instinct should be to say no to both.

You can get pretty far with much simpler techniques.

Should always try to first.