Jake

Building @try_whitney | alaskan, @v0 ambassador | prev @cloudcruise_app (yc w24), @microsoft

[Jake @jakekang]

I recently became a @v0 ambassador!

As a kid, I was addicted to this one airplane shooter game and really wanted to play it again. With the free credits, I figured the "free0" challenge was the perfect opportunity.

I ended up creating: https://t.co/XRPlmPnu3u

I had a ton of fun experimenting with v0 and never imagined actually winning anything.

It’s clear that an engineer’s responsibility will grow beyond just coding and personally, I’m investing a lot of time into learning UX design. With AI tooling only getting better, the only limitation will be our imagination.

I have a few more games to create, but really looking forward to sharing my designs in the future!

[Nick @nickmecklenburg]

one of my favorite parts of leading the post-training forum at south park commons is the speaker series, where we learn from teams exploring interesting directions in industry! next up is Trajectory (https://t.co/wSDeJl1g3S) in SF -- excited for it. https://t.co/7eHHgpch1z

[Jake @jakekang]

@levelsio cracked and battered is pretty good across the street

[Jake @jakekang]

@JayaGup10 where is their spend largest and which groups are responsible? Are people using Opus 4.8 to draft emails?

hopefully they have monitoring for the low hanging fruit

[Jake @jakekang]

@gdb even just 3 months ago. it all changes so fast

[Jake @jakekang]

@satyanadella absolutely! I believe the future will be companies owning their own ai. their data, sops, workflows and the loop that enables continual improvement of their agents.

[Jake @jakekang]

@AnthropicAI Wow this is insane

[Jake @jakekang]

@ekzhang1 and this is why coding will never be solved

[Jake @jakekang]

@vatsalbajaj @AnkurKJain @nickmecklenburg @try_whitney @southpkcommons Excited for what's to come!

[Vatsal Bajaj @vatsalbajaj]

Today, @AnkurKJain, @jakekang, @nickmecklenburg , and I are excited to launch Whitney AI (@try_whitney)

My best friends (and now co-founders) and I all quit our jobs and dove into the unknowns of building a startup. We started our journey this January @southpkcommons spent months wading through the idea maze.

Our team comes from Microsoft Superintelligence, Netflix, Google X and YC startups.

At Whitney AI, we see a massive missed opportunity. Today’s AI-native enterprises have squeezed everything they can out of frontier models via better prompts, better scaffolding and better context. But if every company is using the same general purpose models, where does that leave your competitive edge?

Your real edge is in the assets that your competitors can’t copy: proprietary data, evaluations and production traces. And yet, most companies fail to harness them.

Whitney AI fixes that. We embed with your team to build self-improving custom models and evaluations on your proprietary data, so the intelligence you walk away with is owned by you and compounds over time. We help you build your AI moat.

Here’s a picture of my team getting ourselves flowers after a hard day at work. When we deliver for you, the next bouquet's yours.

[Jake @jakekang]

@vatsalbajaj depends on the company

[Jake @jakekang]

@levelsio my quince bag has been treating me great so far. they have the hinomoto wheels at a much better price than more expensive brands

[Jake @jakekang]

The narrative I hear is that token costs are rising at a tremendous rate. But every AI leader I talk to say cost is not an issue for them.

Everyone is focused on growth, from seed to enterprise. As long as the workflow value you get out of AI is greater than the cost, then companies should remain focused on squeezing out value.

Sure, when we start to reach a plateau, then we can start optimizing for cost and/or latency. But that doesn't seem to be the case today or in the next 6 months.

[Jake @jakekang]

Model routers are a bit dubious.

Give the providers your prod traffic for something that can be built in a day? protecting your data / ai is worth so much more than having to trial and error the best models for your agents

[Jake @jakekang]

@saranormous I've done my part by getting my mom in rural alaska to use chatgpt

[Jake @jakekang]

@harvey @FireworksAI_HQ SFT is a bit outdated, would be curious to see more modern techniques applied. I would imagine GLM 5.1 + Opus 4.7 would be worse off then

[Jake @jakekang]

@samdblond what about carbon offsets?

[Jake @jakekang]

@levie How are they measuring productivity via ai spend? at some point the productivity gains will vastly be larger than ai spend, at which whatever billions they were spending will be worth it.

[Jake @jakekang]

@weill @CreatorML_ I imagine you had some classical ML approach to solving this?

although an LLM-based approach can be interesting if you did an artificial societies type approach for a youtube video (packaging + actual content of video, timestamps, etc.)

[Jake @jakekang]

@ClementDelangue Foundation models both overserve and underserve many use cases. But post-training should only be done if:

- you are interested in learning
- you have achieved PMF

pre-pmf, there is nothing to optimize for

[Applied Compute @appliedcompute]

Your data is your edge, but only if your AI is built on it. Rent a generic model and so can your competitor. The companies with an edge are deploying custom models that they own and improve over time.

Our co-founder @rhythmrg recently stopped by @southpkcommons to share how companies are owning their intelligence with Applied Compute.