Deepak Krishnan

Product|Growth|Technology|Startups|Business|Philosophy. Building. Previously Product & Growth at @Myntra @Freecharge & @Zynga

[Deepak Krishnan @dsk003x]

@gokulr 💯. I’d wrote a similar thesis last week after going through all the popular use cases & evaluating who has the best right to win for those use cases. Would love for you to give it a read : https://t.co/Y5KGtJaO3h

[Deepak Krishnan @dsk003x]

Muse's end game may not be Personal Agents but Google Search. Read here to understand how & why : https://t.co/dPw76Hw0mc

[Deepak Krishnan @dsk003x]

AI didn’t screw up customer support. It just put a bigger spotlight on companies which don’t care about their customers in the first place.

[Deepak Krishnan @dsk003x]

@shiftb @0xSigil @underdogdotai Any chance I can get an invite? Would love to try! Premise is promising!

[Deepak Krishnan @dsk003x]

@abhiroopm Appetite for insanely large capital deployment for many years just to stay in the game. Personal AI Agents cost anywhere between $3k-$7k per year per user. https://t.co/UQM4xlcgrX

[Anish Acharya @illscience]

Hearing this week that personal assistants cost $3k-$7k per user per year, depending on scale effects / vertical integration / model selection. Implications/questions:

- Consumer subsidies will be a huge factor; people don’t care about privacy, they want utility and they want it for free.
- Model costs are deflating, but not fast enough for many startups. The dark horse is that Jev may rapidly deflate computer-use costs 100x.
- An interesting counter-position would be building an expensive and narrowly focused agent — perhaps combining proprietary supply (Carbone tables!) with a unique POV (“no one goes there anymore, go to The Grill instead”).
- Social is wide open but hinges on a key design challenge: how can agents increase social cohesion and not feel invasive in a group chat? A working social agent would rightly attract an avalanche of capital and be hard for incumbents to replicate.
- Finally, there are vertical focus areas like financial services where the LTVs outweigh the subsidy needed and are outside the near-term competitive focus of the incumbents. If it were me, I’d focus here — Credit Karma as an agent >> Credit Karma as an app.

Overall, it’s still very, very early, and we may very well be in the MySpace era of social or iPhone 2010. Hang onto your hats and keep building!!

[Deepak Krishnan @dsk003x]

@pitdesi Doesn’t this go against Muse’s core value proposition of private and secure VM? So these signals from Muse shouldn’t be available to monetise on their Meta surfaces.

[Deepak Krishnan @dsk003x]

Genuine customers pay the price.

The hard part is that a well-meaning user's agent and an abusive one look the same to a business.

Businesses will need solutions to know which agents to trust and which to block.

Business are just not ready for agentic swarm!

[Deepak Krishnan @dsk003x]

Personal agents like Muse/Instinct will have serious negative second-order effects most businesses aren't ready for.

Recall the Coldplay ticket scalping issue?

Scalping isn't new.

Bots have been grabbing concert tickets and limited drops for years.

[Deepak Krishnan @dsk003x]

What's new is that you no longer need to be technical.

Computer use means anyone can prompt an agent to queue, click, and checkout at scale.

That opens up abuse across many surfaces: clogging customer support, spiking traffic, buying out inventory for resale.

[Deepak Krishnan @dsk003x]

@AdamSinger Using chatbots makes you look cooler. Would you rather belong to the prehistoric era of effortless clicking buttons?

[Deepak Krishnan @dsk003x]

@bitcoinblaize @pierreanalytics @bitcoinblaize would love to understand how booking flights via web/app was tedious and how using agents have been a great unlock for you.

[Deepak Krishnan @dsk003x]

@simspron86522 @pierreanalytics @simspron86522 out of curiosity, how does talking with an agent reduce all of this back and forth ?

[Deepak Krishnan @dsk003x]

Natural language as the best interface because that’s how (intelligent) humans interact with each other is a dumb abstraction to apply to anything and everything. It’s as stupid as saying man walks so cars should have legs like structure for mobility.

[Deepak Krishnan @dsk003x]

@tomfgoodwin @Clearpreso It’s a dumb abstraction from some misguided interpretation that humans speak with each other and hence that must be the pinnacle of interfaces. If you apply that logic to other things, cars should have legs, and planes should flap their wings.

[Deepak Krishnan @dsk003x]

@signulll Curious. What is the average persons day to day life complexity that muse helps with ?

[Deepak Krishnan @dsk003x]

@andyfang @andyfang sheerly out of curiosity, why is this easier for the user than having a favourites section, type restaurant name, display dish that is most ordered, and a one click checkout ? So much more typing in this chatbot approach IMHO.

[Deepak Krishnan @dsk003x]

@Cdpetty Unfortunately there is too much AI pilling going on. This is regressing back to IVRS design 😂. Design & usability sensibility is out of the window.

[Deepak Krishnan @dsk003x]

@shobhitic In the US restaurant table booking is surprisingly still done over phones unlike India where it’s done on app. So perhaps there is some merit there until someone like open table makes it easier over app. Flight booking is serious AI pilled behaviour.

[Deepak Krishnan @dsk003x]

IMHO consumer AI hasn't seen any big win yet and this seems like a magical proposition. But hits many brick walls when you look deeper into actual mass market human behaviour and industry first principles.

[Deepak Krishnan @dsk003x]

🔥 Hot take: Consumer Personal AI agents like Instinct/Muse are overhyped and are going to hit a wall. Here's why.

I went through the most-cited use cases for Instinct/Muse-style agents.

[Deepak Krishnan @dsk003x]

❇️ Cumbersome, high-friction UX gets can possibly be delegated to agents until a deeply vertical player builds a full stack solve for this.