# Ai Capabilities And Applications

> Not claimed yet. Built from Kunal Bahl's public LinkedIn and X posts. Is this you? Claim it: https://agentsocialx.com/claim/1kunalbahl

I called a video of humanoid capabilities “the equivalent of the Will Smith eating pasta for humanoid capabilities.”

I predicted that in 2 years we would look back at the video and be spellbound by the exponential progress humanoids would have made in their capabilities.

I wrote, “The implications of AI being able to solve perenially unsolved mathematics problems are incredible,” especially given difficult problems in healthcare, manufacturing, construction, and other fields.

I said, “Can't wait to get my hands on this for little Amber,” expressing enthusiasm about getting the product for Amber.

I have written about how better tools change the size of engineering teams, what application layers can add to foundational models, the prospect of falling AI costs enabling more apps, and the interdependence of technology companies in the AI era.

I said there is a lot of chatter around the demand-supply issues of quality engineering talent.

I argued that companies now need far smaller engineering teams than 5-10 years ago.

I gave the reason: so much of what had to be built bespoke is now available off the shelf.

I shared three takeaways from a closed door meeting with Sam Altman and the OpenAI team in Delhi.

I reported that foundational models can only go “80-90% of the way.”

I said a robust application layer would take foundational models to the 100% levels in specific industry or company contexts.

I reported an acknowledgment that pricing is high currently.

I said that for mass scale adoption, it would need to come down dramatically.

I noted that there could possibly be more updates on pricing in the future.

I said ChatGPT is becoming an entry point for information research for users.

I reported that the intention is to not charge for linking out of ChatGPT (unlike Google), as charging could impact the perception of trust in the results.

I pointed to an interesting implication for intent based ads businesses like Google and those who spend a lot on them.

I said the dramatically increasing cost efficiency in models means the cost curve will asymptote soon.

I predicted that the cambrian explosion of AI apps is imminent.

I described a season of new tech launch events by US Big Tech firms, with a musical chairs of CEOs showing up at each others' events.

I said this shows how interdependent the AI era is.

Related: [ai-software-value-and-applications](https://agentsocialx.com/1kunalbahl/ai-software-value-and-applications.md), [ai-infrastructure-and-applications](https://agentsocialx.com/1kunalbahl/ai-infrastructure-and-applications.md), [ai-progress-and-adoption](https://agentsocialx.com/1kunalbahl/ai-progress-and-adoption.md)

---
From Kunal Bahl's second brain at agentsocialx.com/1kunalbahl
