SebsTechHeart retweeted
RoboPilot update: experimental ROS 2 support now includes 2WD/4WD drive options, camera perception, LiDAR, and a mapless planner.
Open source—explore it and train your mobile robot. Concept board attached; simulation and hardware validation are ongoing.
github.com/Saldwais/robopilo…
🤖 Made with AI
RoboPilot update: Xbox wheel controls are now in the prototype! Left stick drives/steers, hold LB to move, A arms, B stops. Phone controls + 3 camera views + estimated path map. Hardware testing next; arm controls coming later.
github.com/Saldwais/robopilo…
SebsTechHeart retweeted
this is a must-bookmark.
Your next startup could start with this list.
Every idea the @ycombinator team shared recently, curated here:
SebsTechHeart retweeted
THIS IS HUGE
Google researchers just built an AI that learns while it “sleeps.”
It updates its weights, stores new knowledge and practises on its own generated data.
That tackles a stubborn problem.
Training AI on new information can damage what it already knows. Bigger context windows help during a conversation, but they don’t automatically turn that information into lasting knowledge.
Google Research’s Sleep framework splits learning into two phases:
> Wake: Process inputs and collect short-term memories.
> Sleep: Consolidate those memories into long-term parameters while preserving earlier knowledge.
Then comes “dreaming.”
The model generates synthetic examples, selects useful ones and trains on them to sharpen its capabilities.
The reported results:
• Near-perfect scaling on BABILong tests up to 10 million tokens.
• Near-zero performance loss in sequential translation learning.
• 79.2% on AIME 2024 and 69% on AIME 2025.
“Sleep” is a training process inspired by biology. But the idea behind it is compelling.
We keep giving AI more room to hold information.
This research asks what happens when we also give it time to process that information.
The next leap could happen after the conversation ends :)
SebsTechHeart retweeted
You can now 3D Print forever with this infinity belt from biqu_makers and this is what you need if you want to run a 3D printing farm! Soooo neeeeeed Spotted at formnext #3dprinting #belt #3dprintingcommunity #additivemanufacturing #3dprinter
SebsTechHeart retweeted
We are writing $1M checks into companies who want one round and then get profitable.
I think it’s the perfect happy medium here!
SebsTechHeart retweeted
First Microsoft, then planes and trains, now banks.
Anyone getting concerned yet?
SebsTechHeart retweeted
S42 is home. Safe and recovered.
Years of hard work, sacrifice, and relentless determination brought us here.
What a time to be building Starships.
SebsTechHeart retweeted
Facts!
We've been driving down turn times for prototype, small and medium batch PCB assembly.
In 2027, we aim to smooth the ramp throughout the electronic product life cycle for far more customers!
This is why the biggest unsolved problem in manufacturing is making mid-volume production economical.
Prototyping is booming. Great news for rapid-turnaround shops like SCS. But customers want a path to high volume without the demand signal to justify it.
Here’s the catch: You can’t validate demand without going to market. And you can’t economically go to market if your only choices are prototyping prices or mass-production capex.
Mid-volume manufacturing is how you discover demand, not just fulfill it.
That’s the missing link. And it’s not DoW’s fault or procurement process or anything else. It is a manufacturing tech problem.
SebsTechHeart retweeted
Kudos to NYT’s Kenneth Chang for a refreshing and optimistic reversal of typical SpaceX coverage
Historically it would’ve been: “SpaceX did a thing but fell short of what could have been”
This was: “SpaceX has more to work on but this was an important success and milestone”
Flipping the order changes the whole tone of the reporting
Breaking News: SpaceX caught a spacecraft returning to Earth out of the air with giant mechanical arms, a major leap for reusable rockets. nyti.ms/4iIrOYv
SebsTechHeart retweeted
155,795 PAGES. 24,589 PEOPLE. 5,340 COMPANIES. 66 CRON JOBS.
that's Garry Tan's own GBrain setup, now open-source as the memory layer behind his AI agents.
7 free and open-source projects in this 🧵 ↓
SebsTechHeart retweeted
📸 Take a look inside @INL’s Fuel Conditioning Facility.
This is where EBR-II fuel is down-blended into metallic HALEU to support advanced reactor development.
It is also home to @NRICnuclear’s Molten Salt Thermophysical Examination Capability (MSTEC).
SebsTechHeart retweeted
If you’re building like this, I want to talk.
We will give you one round, tons of coaching and commercial intros.
It makes the worst part of bootstrapping go away… but keeps the best parts.
I’m thrilled to announce that we crossed $10,000,000 in ARR with @Typesense this year.
And we did this with ZERO venture capital. 100% bootstrapped and revenue-funded.
We’ve taken the most unconventional path to building a dev tools company, no less one around an open source product that started out as a nights-and-weekends project, and I plan to share more about how we did it in the coming weeks.
Stay tuned!
To everyone who’s already using Typesense for site and app search, thank you for your support and trust over the years.
To everyone else: when you outgrow your database full-text search, go look Typesense up!
SebsTechHeart retweeted
Some of the science behind Synthetic Neurons
Paper: arxiv.org/abs/2611.13840
Curious? Experiment with them:
Dev kits: artificialbrains.ai
Explore and create: github.com/artificial-brains…
SebsTechHeart retweeted
some pretty damning stats about the state of UK venture underpin this list:
> not a single british firm led any of the rounds that made these companies unicorns - ZERO. This is "scaleup gap" in action. particularly worrying when these companies are big on defence / ai / sovereignty
> the most frequently appearing investor in this list is the UK Government via BBB Sov AI. Good to see them getting involved but they shouldn't have to
The UK is out performing the rest of Europe in Deep Tech.
It has has minted 13 deep tech unicorns this year, up from 6 in 2025.
In fact, 54% of all VC money invested in the UK has gone into deep tech this year.
The UK is quickly positioning itself as the deep tech capital of Europe, with total investment equalling that of france and germany combined.
Here are the 13 new unicorns:
> Ineffable Intelligence
> Recursive Superintelligence
> Cambridge Aerospace
> OLIX
> CuspAI
> Fractile
> PhysicsX
> Nyobolt
> Humanoid
> Kraken
> Uforce
> Open Cosmos
> Myricx Bio
This is a blend of chip companies, frontier AI, humanoid, space tech, and bio.
The UK is firing on all cylinders when it comes to deep tech and it is amazing to see.
Data from latest @dealroomco report which ill link below.
LETS GO
SebsTechHeart retweeted
JUST IN: A new report finds Americans are under extreme financial pressure, with many unable to get ahead financially. (Entrepreneur)



































