lamroger

We are so back

College might be a good starting point since you're young

Philly one is good too

Love it!

What if you just ask chatgpt directly?

Josh started with 50, ended with 20. Jane started with 100, ended with 130.

although i could see it being helpful to keep track of stages like when people start buying in and calculating at the end.

chatgpt might deal with state well in the context but not 100% reliable.

but maybe?

Snap, im guessing a lot of media is still recovering from the strike. I know a lot of vfx people are still looking. Gaming and media is getting hit hard but that's not LA specific but a big sector.

Man the LA market is getting squeezed

> If your job isn’t enough to satisfy your curiosities, look to contribute to FOSS.

Find ways to challenge yourself that also compounds.

I haven't worked in big tech but I imagine their tooling is really good for this particular stack. Internal expertise too for running into big scale edge cases.

ChatGPT formality as another flavor for code switching. I could see it. Good or bad? I think it's good if they're testing how to be persuasive and if it works.

Forgot about Chef... Good times...

One of his predictions is FemTech being bigger. More data to collect and mine.

I just hope these startups have better security than 23andme

Which round matters too

I'd think it's really good at tokenization to word. That's exactly what it's trained to do.

  Location: Los Angeles, CA
  Remote: Remote / Hybrid
  Willing to relocate: No but can do 1 week per month in the Bay
  Technologies: Generalist but strength in Backend/Infra
  Résumé/CV: https://www.linkedin.com/in/lam-roger/
  Email: Best way is to LI connect / message

Worked as a software developer for 9 years - mostly startups and developed a knack for figuring things out.

Got laid off a few months ago and spent a few months catching up on LLMs and seeing if a few projects stuck (LLMOps). Now applying for jobs and new personal goal is to be a core contributor to Pytorch.

Would love to find something local since I've been remote even before COVID and miss the in-person collaboration.

Everyone that would have bought the album already has it

I'm also in LA and having a similar experience. Not much of an AI startup boom. I was hoping on finding something of a hybrid situation after being tired of fully remote.

I wonder if using APIs was more of a first to market move

I like that Andrew brought up the idea of containment. I've lived in the major cities he described and it's the same pattern. It's invisible to the well off neighborhoods. The poor neighborhoods have to live with it. The fact that it's always been there - just hidden.

Re: tech - it def disrupted the fabric of the city. As quickly as it came (gentrification, raising rents) and as quickly as it left (residents losing adjacent jobs), it's been bad for the people who had to live through it without high paying tech jobs.

I liked that Andrew touched on the opioid epidemic and how it's spiraled. Heavy policing historically disenfranchised poor minorities so that's not the way to go. Progressive policies were tried and the people pushed back.

I wish there was a single solution. It feels like we fell off a cliff and the only way up is to crawl back up.

Selling ads

i love how they used the word advanced. not another regular AI team.

Automated alerts - SRE as a service. Anomaly detection with some good suggestions. SRE intern as a service? Something like that - notifying you of something vs human requesting.

Make me think that a push model would be awesome. Would that be an agent?

Spy stuff

DD learned from NR's mistakes. But they're both expensive now.

dope yea that's awesome!

Cellular automata bridged math and CS for me and it's beautifully chaotic. It's the same backwards with deep learning. Chaos to structure in classifying numbers or "doing" math.

It's the unexpected that's exciting.

The guardrails are cool!

I think more details of where the data goes and when it goes from few-shot to fine-tune will be helpful.

Philly replaced a burnt bridge pretty quickly. Not sure how different the sizes are but hoping for a speedy fix.

https://www.businessinsider.com/philadelphia-turns-heads-spe...

I wish we had some insight on offshoring numbers

The general consensus imo is that fine-tuning is more for tone and style vs accuracy. People use vector DBs to grab relevant data to throw into the prompt and call it Retrieval Augmented Generation.

From what this seems to do is host multiple deltas fine-tunings and hot swap as needed. Incredible optimization. It's like going from AMIs to ECS or Kubernetes.