Pinned Tweet
Changed my entire career path for this dream.
One of the happiest days of my life.
From a render to reality, Starship's journey inspires me every day. Soon, I hope to be a part of it.
Per aspera ad astra 🚀✨
One of the happiest days of my life.
From a render to reality, Starship's journey inspires me every day. Soon, I hope to be a part of it.
Per aspera ad astra 🚀✨
23
38
526
72,194
Eric Magvas 🇺🇸 🚀 retweeted
Don't forget to join the debate tonight - we've got some people from outside the space community to discuss some important aspects of space colonisation. The propositions:
1. 'If microbial life is found on a planet, we should not land on that planet, but leave it to its own development.'
2. 'If we never find life beyond Earth, the case for atheism becomes weaker.'
3. 'Science fiction and long-term scenarios are essential for understanding our future.
4. 'We should colonise space as soon as we are able.'
x.com/i/spaces/1RKjp…
1. 'If microbial life is found on a planet, we should not land on that planet, but leave it to its own development.'
2. 'If we never find life beyond Earth, the case for atheism becomes weaker.'
3. 'Science fiction and long-term scenarios are essential for understanding our future.
4. 'We should colonise space as soon as we are able.'
x.com/i/spaces/1RKjp…
10
4
39
6,660
Eric Magvas 🇺🇸 🚀 retweeted
This seems half right. We'll find 100,000's of vulns, which over time will leave no choice *but* to migrate to systems where perfection is not only possible, but provable.
Just going to preregister here that I believe the computer security overhang is not a few dozen bugs that get patched and then we're good, there are 100,000s of vulns that are extremely discoverable with infinite free tokens and we're probably going to have to move away from systems that presuppose that perfection is possible.
0
1
15
5,449
Eric Magvas 🇺🇸 🚀 retweeted
@aphysicist Interesting
"we're going to be overwhelmed with slop" people, 6 months ago.
1
1
10
1,556
Eric Magvas 🇺🇸 🚀 retweeted
@IterIntellectus Can we improve the restorative functions of sleep without sleeping longer?
The next step is measuring and controlling glymphatic clearance.
If we can do that, we may be able to make fragmented sleep more restorative.
The next step is measuring and controlling glymphatic clearance.
If we can do that, we may be able to make fragmented sleep more restorative.
2
1
12
521
Eric Magvas 🇺🇸 🚀 retweeted
Join us!
Peter Hague and I are organising a debate between scientists (@peterrhague), engineers (@TheVolqaGuy, @JoelSercel), philosophers, and theologians (@meharmsen) here on X. The Theme is: 'The place of Humanity in space'.
Set your Calendars: Thursday 21.30 London, New York 16.30, Los rs can join the debate in the comments.
1. 'If microbial life is found on a planet, we should not land on that planet, but leave it to its own development.'
2. 'If we never find life beyond Earth, the case for atheism becomes weaker.'
3. 'Science fiction and long-term scenarios are essential for understanding our future.
And last but not least:
4. 'We should colonise space as soon as we are able.'
5
7
55
2,376
Eric Magvas 🇺🇸 🚀 retweeted
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com
878
2,614
19,734
4,461,101
Eric Magvas 🇺🇸 🚀 retweeted
Check out the upgraded Mars city environment at domes.caseyhandmer.com. Absolutely crazy how much better I can tell the story of open spaces with tensile columns vs just seven years ago, when I managed one crappy render. This project has really pushed the limits of procedural s city depicted would take about 12 hours to walk across. The render rewards a capable GPU but it also works on mobile. Reply in thread with screenshots of your favorite features and/or bugs.
I finally worked out to mathematically express this surface, which allowed me to render my vision for large pressurized living areas on Mars.
Light - ✅
Tensile - ✅
Extensible - ✅
Open area - ✅
Not a dome - ✅
Transparent - ✅
Self-supporting - ✅
Simple foundation - ✅
45
28
354
22,804
Eric Magvas 🇺🇸 🚀 retweeted
Twenty years ago, I graduated from @pepperdine. It’s shaped my life in many ways, including being where I met my wife, Betsy. My time on campus, traveling overseas and studying the great books taught me how to communicate effectively and build with a focus on mission.
Today I’m proud to support Pepperdine’s work to open doors for students to lives of purpose, service and leadership. Thank you to Jacqueline for giving me the chance to share some of what I’ve learned at Pepperdine and since:
pepperdine.edu/newsroom/artic…
Today I’m proud to support Pepperdine’s work to open doors for students to lives of purpose, service and leadership. Thank you to Jacqueline for giving me the chance to share some of what I’ve learned at Pepperdine and since:
pepperdine.edu/newsroom/artic…
5
6
51
2,425
Eric Magvas 🇺🇸 🚀 retweeted
@masayoshisson @3F_Research 6 years seems conservative tbh.
I think with disaggregated inference you will see Blackwells have useful lives over 10 years.
Will be able to put the latest and greatest SRAM accelerator in front of a B200 in 5 years and make it sing.
I think with disaggregated inference you will see Blackwells have useful lives over 10 years.
Will be able to put the latest and greatest SRAM accelerator in front of a B200 in 5 years and make it sing.
18
34
603
169,702
Eric Magvas 🇺🇸 🚀 retweeted
Tower view of the first Super Heavy booster catch
1,271
6,312
41,490
2,253,346
Eric Magvas 🇺🇸 🚀 retweeted
HIRING! SO MUCH HIRING AT POSEIDON🚨✈️
SENIOR FLIGHT SOFTWARE ENGINEER
SENIOR AVIONICS ENGINEER
SENIOR GNC ENGINEER
SENIOR SIMULATION, TEST, AND VALIDATION ENGINEER
SENIOR MECHANICAL STRUCTURES ENGINEER
FLIGHT SOFTWARE ENGINEER
AVIONICS ENGINEER
GNC ENGINEER
SIMULATION, TEST, AND VALIDATION ENGINEER
AND 12 MORE
SENIOR FLIGHT SOFTWARE ENGINEER
SENIOR AVIONICS ENGINEER
SENIOR GNC ENGINEER
SENIOR SIMULATION, TEST, AND VALIDATION ENGINEER
SENIOR MECHANICAL STRUCTURES ENGINEER
FLIGHT SOFTWARE ENGINEER
AVIONICS ENGINEER
GNC ENGINEER
SIMULATION, TEST, AND VALIDATION ENGINEER
AND 12 MORE
16
18
238
9,001
Eric Magvas 🇺🇸 🚀 retweeted
The great metals shortage will be exposed.
BREAKING: The U.S. Army is preparing to stress-test America’s manufacturing capacity and supply chains. In early 2027, Army facilities and private manufacturers will compete in the first-ever “Manufacturing Rodeo,” designed to expose bottlenecks in America’s defense industrial base.
22
58
1,101
53,003
Eric Magvas 🇺🇸 🚀 retweeted
After 18 months of work, FlightAtlas.org now has a drawing for every rocket launch in history.
To celebrate I'll post 70 visual launch overviews over the next month, one for every year from 1957 to 2026 (~2 per day).
Hope you guys enjoy!
To celebrate I'll post 70 visual launch overviews over the next month, one for every year from 1957 to 2026 (~2 per day).
Hope you guys enjoy!
12
26
162
16,020
Eric Magvas 🇺🇸 🚀 retweeted
The foundations for an exciting future are being built in Texas. Next up: Terafab → spacex.com/updates#terafab
13
10
101
7,712
Eric Magvas 🇺🇸 🚀 retweeted
we are looking for exceptional software engineers who can help us design and operate vehicles that are reliable enough, performant enough, and simple enough to fly 100,000 times a year.
DM me your resume before applying. job-boards.greenhouse.io/spacex/jobs/85…
DM me your resume before applying. job-boards.greenhouse.io/spacex/jobs/85…
10
14
428
30,230
Eric Magvas 🇺🇸 🚀 retweeted
Super intelligence will be the greatest accelerant of scientific progress in human history and will transform every scientific discipline.
.@mkratsios47: "The thing we're most excited about is applying Super Intelligence to scientific discovery. We fundamentally believe that all scientific domain — whether working in material science, in chemistry, in math, in physics — are all going to be accelerated by the use of SI."
52
132
744
66,355
