Thomas Shehadeh retweeted
Introducing Fern, Mirendil’s first model: pretrained from scratch, with up to 10M tokens of context!
That’s ~0.5M lines of code (roughly Megatron-LM scale) or days of an agent’s work, all in context at once.
Its first version, Fern 0.1, had a short round of post-training 🚢
That’s ~0.5M lines of code (roughly Megatron-LM scale) or days of an agent’s work, all in context at once.
Its first version, Fern 0.1, had a short round of post-training 🚢
133
198
2,340
210,528
when i first met the founders of the residency, i was immediately inspired by the vision of nick and peter and their approach to cultivating young talent - amazing founders continue to emerge from the residencies across SF! thank you for having me
Math is the next frontier for AI training data.
I sat down with @tshehads for a fireside chat at the residency and hearing him make that case to our founders was the highlight.
Thanks for joining us, Thomas.
0
0
3
164
Thomas Shehadeh retweeted
Math is the next frontier for AI training data.
I sat down with @tshehads for a fireside chat at the residency and hearing him make that case to our founders was the highlight.
Thanks for joining us, Thomas.
I sat down with @tshehads for a fireside chat at the residency and hearing him make that case to our founders was the highlight.
Thanks for joining us, Thomas.
0
4
14
348
Thomas Shehadeh retweeted
Introducing SKY-001, the first Engineering World Model.
On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested:
-> 51-125x faster
- > 150-210x cheaper
-> Higher IOU than all frontier models
SKY-001 lets you dream a physical object before you build it.
18 months ofm out of Maluuba's deep learning lab (acq. by @MSRResearch).
The first ‘AI General Engineer’ will be the backbone for Autonomous Manufacturing.
On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested:
-> 51-125x faster
- > 150-210x cheaper
-> Higher IOU than all frontier models
SKY-001 lets you dream a physical object before you build it.
18 months ofm out of Maluuba's deep learning lab (acq. by @MSRResearch).
The first ‘AI General Engineer’ will be the backbone for Autonomous Manufacturing.
824
2,413
16,277
5,002,218
excited to be on this journey w sam ! 100x faster and 150x cheaper than astra... unreal!
Introducing SKY-001, the first Engineering World Model.
On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested:
-> 51-125x faster
- > 150-210x cheaper
-> Higher IOU than all frontier models
SKY-001 lets you dream a physical object before you build it.
18 months ofm out of Maluuba's deep learning lab (acq. by @MSRResearch).
The first ‘AI General Engineer’ will be the backbone for Autonomous Manufacturing.
2
1
7
235
if someone could find the team that build linkedin's anti-scraping system i will write a blank check. its been astounding for almost a decade
0
0
0
36
Thomas Shehadeh retweeted
Congrats to @Ultraroboticsco on the raise! 🚀
Ultra and Pi have been working together to bring our models into real warehouses. We've gone from training ~1 policy per week to 8+, and improved autonomous performance from ~60% to ~95%.
Really cool to see how quickly things can improve when you're iterating on models with real-world robot data.
Ultra and Pi have been working together to bring our models into real warehouses. We've gone from training ~1 policy per week to 8+, and improved autonomous performance from ~60% to ~95%.
Really cool to see how quickly things can improve when you're iterating on models with real-world robot data.
Robotics has a distribution problem. Today, Ultra is announcing $62M in funding to solve it, including a $50M Series A led by @hiFramework.
We believe that deployment is the bottleneck in robotics.
It’s relatively easy to make a robot look impressive in a controlled harder to put a robot into the open world, make it useful, keep it running, earn the trust of the people working alongside it, and do that over and over and over again.
So that’s what we’re focusing on.
3
12
48
3,117
excited to work with this team!
Robotics has a distribution problem. Today, Ultra is announcing $62M in funding to solve it, including a $50M Series A led by @hiFramework.
We believe that deployment is the bottleneck in robotics.
It’s relatively easy to make a robot look impressive in a controlled harder to put a robot into the open world, make it useful, keep it running, earn the trust of the people working alongside it, and do that over and over and over again.
So that’s what we’re focusing on.
5
0
23
1,877
Thomas Shehadeh retweeted
Robotics has a distribution problem. Today, Ultra is announcing $62M in funding to solve it, including a $50M Series A led by @hiFramework.
We believe that deployment is the bottleneck in robotics.
It’s relatively easy to make a robot look impressive in a controlled harder to put a robot into the open world, make it useful, keep it running, earn the trust of the people working alongside it, and do that over and over and over again.
So that’s what we’re focusing on.
We believe that deployment is the bottleneck in robotics.
It’s relatively easy to make a robot look impressive in a controlled harder to put a robot into the open world, make it useful, keep it running, earn the trust of the people working alongside it, and do that over and over and over again.
So that’s what we’re focusing on.
104
95
717
229,963
Thomas Shehadeh retweeted
it's finally here
my ranking of all of the ai assistant tools i've tried over the past six months
let me know if you disagree
my ranking of all of the ai assistant tools i've tried over the past six months
let me know if you disagree
137
63
1,420
104,906
Thomas Shehadeh retweeted
What does a cancer vaccine founder have in common with a factory entrepreneur, a five-time founder, the inventor of Ring, the former CEO of Twitter and elite college athletes?
Put them in a room with hundreds of Michigan students and patterns emerge. Five years ago, arbitrage started the Venture Summit with the @ZellLurie Institute to bring students into conversations with leaders across very different fields.
As H01 founder @davefontenot put it at the 2024 Summit: a Silicon Valley tour of duty is essential but the next exceptional founder, idea and capital can come from anywhere.
Founders choose venture firms. Institutional investors decide which firms to entrust with their capital. We started with that capital chain: @UMich's Dan Feder, @openfx_ founder @prabhakar2reddy and M13 co-founder @CarterReum explored how exceptional people recognize one another before the outcomes make the answer easy.
M13 partner Sarah Tomolonius spoke with Serova cofounder Michael Bryan about personalized cancer vaccines and moving with startup urgency when the human body, scientific evidence and institutions operate on timelines founders can’t control.
@ring founder Jamie Siminoff and M13 @wchristinec explored the human behind the founder story: what makes someone unreasonable enough to keep going, how fear becomes fuel and what happens when that fuel no longer serves you.
@fredwilson of @usv and former Twitter CEO @dickc joined Latif to trade unfiltered stories from some of venture and tech’s most consequential years: the deals they made, the ones they missed, what they got wrong and sports and tech predictions.
@SchramIAm of @640oxford, @noxmetals founder @zanehengsperger and @cantos' @ameekapadia explored AI’s move into the physical world, how it is changing American manufacturing and the jobs that will emerge around it.
Venture has plenty to learn from elite sports. Michigan coaches Kyle Church, @KBA_GoBlue, and @thacoachmike, athletes @Gobluety1 and @SylaSwords joined Roger to talk about exceptional performance, building winning teams and what student athletes need from institutions to thrive.
Across very different careers, a few ideas kept resurfacing. Speed matters. Mission sustains people through the hard parts. Exceptional people attract exceptional people. And there is no single archetype for success.
Thank you to everyone who has kept building this gathering with us. Excited for next year. Go Blue.
Put them in a room with hundreds of Michigan students and patterns emerge. Five years ago, arbitrage started the Venture Summit with the @ZellLurie Institute to bring students into conversations with leaders across very different fields.
As H01 founder @davefontenot put it at the 2024 Summit: a Silicon Valley tour of duty is essential but the next exceptional founder, idea and capital can come from anywhere.
Founders choose venture firms. Institutional investors decide which firms to entrust with their capital. We started with that capital chain: @UMich's Dan Feder, @openfx_ founder @prabhakar2reddy and M13 co-founder @CarterReum explored how exceptional people recognize one another before the outcomes make the answer easy.
M13 partner Sarah Tomolonius spoke with Serova cofounder Michael Bryan about personalized cancer vaccines and moving with startup urgency when the human body, scientific evidence and institutions operate on timelines founders can’t control.
@ring founder Jamie Siminoff and M13 @wchristinec explored the human behind the founder story: what makes someone unreasonable enough to keep going, how fear becomes fuel and what happens when that fuel no longer serves you.
@fredwilson of @usv and former Twitter CEO @dickc joined Latif to trade unfiltered stories from some of venture and tech’s most consequential years: the deals they made, the ones they missed, what they got wrong and sports and tech predictions.
@SchramIAm of @640oxford, @noxmetals founder @zanehengsperger and @cantos' @ameekapadia explored AI’s move into the physical world, how it is changing American manufacturing and the jobs that will emerge around it.
Venture has plenty to learn from elite sports. Michigan coaches Kyle Church, @KBA_GoBlue, and @thacoachmike, athletes @Gobluety1 and @SylaSwords joined Roger to talk about exceptional performance, building winning teams and what student athletes need from institutions to thrive.
Across very different careers, a few ideas kept resurfacing. Speed matters. Mission sustains people through the hard parts. Exceptional people attract exceptional people. And there is no single archetype for success.
Thank you to everyone who has kept building this gathering with us. Excited for next year. Go Blue.
1
3
5
1,807
no shade but I would like to know how this is different than a sliding window alg
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
0
0
1
58
Thomas Shehadeh retweeted
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
1,555
5,291
40,806
25,674,675
Cool greco-roman undertones in the magazine resurgence
colossus
arena
palladium
vs.
time
life
fortune
colossus
arena
palladium
vs.
time
life
fortune
0
0
0
52
Breaking: Just heard Brad Lightcap left OpenAI to start a data company, realizing that RL is the next horizon for AI and believes that there is long term TAM durability. He reportedly has 8 figures in cARR and is working with the top labs.
1
0
2
694