@elonmusk@elonmusk
Would latency for inference not be easily resolved using embedded starlink comms on all Teslas?
- v1: starlink attached and powered by 12V battery
- Vehicle data center ops begin when plugged in overnight from user response
- Starlink->ECU->Dojo->back
->DTC as backup
We went from 0 to 2,200 paying customers in under a year by following @ycombinator's 15 rules:
1/ Do things that don't scale. Get your first 10 customers by hand.
2/ Launch now, not when it's "ready". A mediocre product in front of real users teaches you more in a week than 6 months of polishing in the dark.
3/ Charge from day one. If nobody will pay, you don't have a startup, you have a hobby.
4/ Talk to users every single day. The roadmap you need is sitting in your customers' heads, and they'll hand it to you for free
5/ Always hunt the 90/10 solution. For almost any feature there's a way to capture 90% of the value with 10% of the effort.
6/ There are only two real jobs: write code and talk to users. Everything else (conferences, press, VC coffees, corp dev calls) is fake work.
7/ You pick your customers as much as they pick you. 10 users who love you beat 1,000 who kind of like you.
8/ Growth is an output, not a strategy. Grow before product market fit and all you're buying is churn.
9/ Do less, really well. Pick one or two metrics and judge every task against them.
10/ Know if you're default alive. Paul Graham's question: on current growth and current burn, do you reach profitability before the money runs out?
11/ Don't hire until it hurts. Headcount is not progress, it's burn. Every great startup was embarrassingly small for embarrassingly long.
12/ Momentum is the only real moat in year one. Ship something every week, even something tiny.
13/ Every great startup is badly broken at some point. The game isn't avoiding fires, it's how fast you put them out. Again. And again
14/ Ignore your competitors. Startups die of suicide, not murder. In year one, the only company that can kill yours is your own
15/ Startups rarely die from running out of money. They die because the founders fall out. Brutal honesty with your cofounder is the cheapest insurance you'll ever buy
Good luck !
@elonmusk@elonmusk
Would latency for inference not be easily resolved using embedded starlink comms on all Teslas?
- v1: starlink attached and powered by 12V battery
- Vehicle data center ops begin when plugged in overnight from user response
- Starlink->ECU->Dojo->back
->DTC as backup
The Haven-1 flight article has been painted. Next, key components including the hatch and domed window will be integrated ahead of pressure and load testing in Mojave, CA.
MCP is getting so good!
Most developers think MCP is just another tool calling standard, but that’s just scratching the surface.
Here's what makes MCP powerful (explained with examples):
Unlike simple tool calling, MCP creates a two-way communication between your AI apps and servers.
Let me break down the 6 core primitives that make this possible:
Starting with the client!
A client offers 3 capabilities: sampling, roots, and elicitation.
1️⃣ Sampling
The server can ask the client to generate LLM completions, but the client still controls permissions and safety.
Example:
A travel server asks the LLM to pick the best flight from a list.
2️⃣ Roots
Clients define what files the server can access. Secure, sandboxed, and scoped.
Example:
A server for booking travel may be given access to a specific directory, from which it can read a user’s calendar.
3️⃣ Elicitations
Servers can request user input mid-task, in a structured way.
Example:
A server booking travel may ask for the user’s preferences on airplane seats, room type or their contact number to finalise a booking.
Moving on, let's talk about the MCP server now.
Serve also exposes 3 capabilities: tools, resources, and prompts
4️⃣ Tools
Controlled by the model, tools are functions that do things: write to DBs, trigger logic, send emails, etc.
Examples:
- search flights
- send messages
- create calendar events
5️⃣ Resources
Controlled by the app, resources are the passive, read-only data like files, calendar, KBs, APIs.
Examples:
- retrieve docs
- read calendars
- access knowledge bases
6️⃣ Prompts
Controlled by the user, prompts are pre-built instruction templates that guide how the LLM uses tools/resources.
Examples:
- plan a vacation
- draft an email
- summarize my meetings
👉 If you're interested in learning more about MCP, let me know in the comments.
I've created a FREE illustrated guidebook with over 10 projects, along with a YouTube playlist for those who prefer videos.
_____
That's a wrap!
If you found it insightful, reshare with your network.
Find me → @akshay_pachaar ✔️
For more insights and tutorials on LLMs, AI Agents, and Machine Learning
@RepMarkWalker@realDonaldTrump Heartiest congrats on your nomination as Ambassador at Large, @MarkWalkerNC! There’s truly no better individual for this role—your unwavering, faith-driven advocacy for people and everyone counts stance is next to none.
#googlecloudready
Am excited to have com[leted Google Cloud Essentials in qwiklabs from spinning machines to deploying on kubernetes and load balancing. Phew!!!https://t.co/tMqyS6rKxC
@WhoIsHLM When a student owes any fees with NCAT, the entire life of the student comes to a stand still. Worse being out of state. Holds on your account, can't register, can't graduate, can't attend classes; international student @ huge risk of deporting. Refund monies now. Unlawful words
@neiltyson Hi Neil, can you write letters to teach me Finite Difference Methods and Advanced numerical analysis. Am struggling in Math Class. You make hard things easy.
DSC 2019 Summit at Google Sunnyvale and Mountain View campus was lit. Silicon valley future developers met here to change the world!!!
Oh, wait a minutes story is unfolding... https://t.co/JTY4h6MpZ3
@mclaurenseries I have a question: Whatever happened to initiating a conversation with someone who is passionate about what you do among other things? aka compatibility?..... Is staying single a statistically significant conclusion. Evolution evokes survival from best traits, why is that absent?