Chalam PVS

Founder @SiftlyAI · YC W22
Enabling brands win Customers from AI searches!

[Chalam PVS @Chalam_PVS]

Search shift:
Past (Google): Ppl srch → Ppl use info
Now (ChatGPT): AI srch → Ppl use info
Future (agents): AI srch → AI use info

Present web content is built 4 humans - keywords, slick UI

For AI: visuals = noise; Semantic context > keyword density

Rules flipped
#ChatGPT

[Rajoshi @rajoshighosh]

We have another "Rebuild with Jev" hackathon this afternoon. If you're in SF, join us today from 1-4pm.

30 mins on Jev architectures/ examples - what works well, how to use multiple models to get the best outcomes + 60 mins to try jev on some of your most expensive workflows - so you can jevvify them effectively.

We have a few spots open. Join us!

https://t.co/agDbhRszD0

cc @typesafeai

[Chalam PVS @Chalam_PVS]

Jev feels like a scikit-learn moment for AI agents

[Chalam PVS @Chalam_PVS]

new ai benchmark just dropped

[Alexandr Wang @alexandr_wang]

state of the art motherfuckers

[Chalam PVS @Chalam_PVS]

@shahidh_k @PromptQL the model is made available on @PromptQL within 4 hrs or announcement? 🤯 Insane speed 🚀

[Chalam PVS @Chalam_PVS]

@rajoshighosh Well, technically @tanmaigo won 🙄

[Chalam PVS @Chalam_PVS]

@marty_kausas Could growing reach explain some of this? higher engagement may attract a larger share of spam and low effort comments

[Chalam PVS @Chalam_PVS]

This feels like a prisoner’s dilemma. If China slows down, the US gains by racing ahead. If China keeps racing, the US races to avoid falling behind. China faces the same logic.

My guess: both keep racing-that’s the Nash equilibrium, even if both would benefit from slowing down together. How do we make restraint worthwhile when neither side trusts the other to slow down?

[Chalam PVS @Chalam_PVS]

Dario’s US-China argument feels like a prisoner’s dilemma from game theory

If China slows down, the US has an incentive to race ahead and gain an advantage. If China keeps racing, the US has an incentive to keep up. Either way, the US is incentivized to keep racing. By symmetry, China faces the same logic

My guess: both keep racing. Neither wants to slow down alone - that’s the Nash equilibrium, even if both would benefit from slowing down together

Escaping that equilibrium means making restraint a better choice for at least one side, even while its rival keeps racing.

How do we make that happen?

[Dario Amodei @DarioAmodei]

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.

Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.

You can read the full post here: https://t.co/OGyPb7yaYt

[Chalam PVS @Chalam_PVS]

@mrsnhl @MonacoGTM @OverlayyAI @AnmolChaman All that midnight oil I saw you burning in sf finally paid off. Congrats congrats!

[Chalam PVS @Chalam_PVS]

@tanmaigo Is it fair to say agents require technical skills to set up, while bots are meant to be used without that?

[Chalam PVS @Chalam_PVS]

@tanmaigo woah.. mine is always walking into the exam hall fully prepared for the wrong subject

[Chalam PVS @Chalam_PVS]

@Chandrika633

[Chalam PVS @Chalam_PVS]

@tanmaigo @rajoshighosh Boardy intro: Tanmai, meet Tanmai 😂

[Chalam PVS @Chalam_PVS]

@mrvaibhav98 @SiftlyAI The ex hangover will take some time to wear off

[Chalam PVS @Chalam_PVS]

@marty_kausas Curious, what is the support fee calculated as a pct of? TCV?

[Chalam PVS @Chalam_PVS]

@tanmaigo 😂

[Chalam PVS @Chalam_PVS]

Unintended side effects of consuming so much AI content:

- humans read faster
- attention spans get shorter

every cold message now has to earn attention even faster

[Rajoshi @rajoshighosh]

Our first guests showed up a week early 😅 If you haven't yet signed up, now is the time! Join us next Tuesday at 5pm for Salt bread & Multiplayer AI!

[Chalam PVS @Chalam_PVS]

Sales/GTM teams need multiple people working with one shared agent - not one person working with 100 agents

Multi-agent AI is here. Now it’s time for multiplayer AI

[Chalam PVS @Chalam_PVS]

@anieasyy 😂