Text adaptation of Andrew Kelley's Software Should Work 2026 keynote andrewkelley.me/post/dont-ta...

This is all well and good, but the big question remains: what is NASA trying to accomplish with PROMISE beyond getting a "win"? What is it trying to accomplish, and is it worth drawing down limited stockpiles of Pu-238 and using our spare RTG to do so?

David Burbach 🇺🇸@dburbach.bsky.social · 5d ago

NASA administrator Isaacman's response to @planetarysociety.bsky.social @caseydreier.bsky.social report that flying the earthbound Mars test rover to the moon will likely cost a billion. Isaacman doesn't actually disagree or cite any inaccuracies, one notes

THIS IS WHAT PEOPLE POWER ACCOMPLISHES! Thank you thank you thank you to everyone who wrote comments, contacted Congress, and yelled! This is OUR WIN! @standupforscience.net www.politico.com/news/2026/08...

Senate finalizes funding patch that would block Trump’s grant overhaul

Language is included to delay the president’s move to put political appointees in charge of grant approval.

politico.com

Did JPL propose a total cost for this, and how much was it?

Dwayne Day@kennen11.bsky.social · 5d ago

The NASA administrator posted this to Twitter. Although I think there is a lot to comment on here, I think the most important point is that he provided a useful marker. 20% of the low-end Planetary Society estimate is ~$100 million. He has now said he will not pursue the mission if it costs >$100M.

Omfg it's literally a Burning Man fever dream.

Dr Cassandra Steer@cassandrasteer.bsky.social · 5d ago

This might be the dumbest space tech prototype ever given approval! This is my rant about Reflect Orbital, and how easy it might be to stop them trying to shine the sun where it don’t shine. substack.com/home/post/p-...

LLMs basically generate bullshit. The official religion of C-Suites is bullshit. Thanks to financialization, corporate performance is completely de-coupled from actually providing goods or services. Put it together and what have you got? Yeah.

By Cory Doctorow (GPG 0xBF3D9110957E5F4C)@doctorow.pluralistic.net · 5d ago

Neoclassical econ assumes rationality. The corollary of, "If you're so smart, why aren't you rich?" is "you're rich, so you must be very smart!" Thus, people assume that if powerful, well-compensated CEOs insist that "AI is changing everything," well then, *AI must be changing everything*. 1/

Omfg it's literally a Burning Man fever dream.

Dr Cassandra Steer@cassandrasteer.bsky.social · 5d ago

This might be the dumbest space tech prototype ever given approval! This is my rant about Reflect Orbital, and how easy it might be to stop them trying to shine the sun where it don’t shine. substack.com/home/post/p-...

We've gone from "shots on goal" to "wins on the board" just like that, huh? Talking past the real risk of failure. This mission, if attempted, will almost certainly result in total failure. The time-to-science and dollar-per-science of a mission that fails are both infinity. Worse than not trying.

Kaynouky@kaynouky.bsky.social · 5d ago

Dreier de la Planetary Society publie un excellent article dans lequel il évalue le coût probable du projet de rover lunaire PROMISE reconverti. Soit entre 0,7 et 1,3 milliard de dollar. Isaacman répond que pas du tout, que si ça coûte plus de ~150 millions ils ne poursuivront pas le projet :)

Isaacman very strongly implied that a "PROMISE" rover already existed, nearly flight ready, with an RTG. "There is another [...] called PROMISE." "Having a nuclear RTG on it [...]" "We've got the hardware." "It exists. [...] It's real." How was the administrator so mistaken about the proposal?

Casey Dreier@caseydreier.bsky.social · 6d ago

I attempted to put upper and lower bounds on what it would cost to repurpose Perseverance's testbed twin and send it to the Moon. It's not the cheap and fast mission people hope it is. www.planetary.org/articles/the...

"A computer can never be held accountable" was supposed to be a warning. But it has been widely interpreted as a strategy.

By Cory Doctorow (GPG 0xBF3D9110957E5F4C)@doctorow.pluralistic.net · 7d ago

I'm not saying that anyone ever sat down and said, "Parents with sick children will pay whatever we charge for a thermometer at 2AM." They didn't have to: this is the kind of thing an automated system can do without any human intervention: www.ftc.gov/system/files... 15/

Africa is smashing all the solar records. The continent installed more grid-connected solar in the first half of 2026 than in all of 2025, and with projects under construction the year could finish at triple the previous record. buff.ly/NirgzLn #ShareGoodNewsToo

African 2026 on-grid solar additions surpass those of 2025 | African Energy

The latest analysis from African Energy Live Data shows despite a slow 2025, solar capacity additions in 2026 have already surpassed each previous year. Falling costs and several recent agreements…

buff.ly

"New space" companies like to claim that (long established) regulations are in the way of their technologies working or business cases closing. One wonders how any space company ever survived at all! When regulations go away (for them), their technologies and business cases still won't work, though.

Lora Kolodny@lorak.bsky.social · last wk.

and now… www.cnbc.com/2026/07/28/s...

On YouTube: About the concept of preemption as the relates to data centers. youtu.be/UdxiyQAcwfM

How Data Centers Avoid Local Planning

YouTube video by City Beautiful

youtu.be

I want this to work, because Swift is awesome. And I think Katalyst has a real / useful business idea. But this is also going about as well as expected, given the TRL and timelines. I'm broadly in favor of legitimate, acknowledged Hail Mary attempts that also de-risk critical technologies.

Fabian Schussler@fabianschussler.bsky.social · last wk.

The Swift rescue mission has some difficulties 😱 ☄️🔭 arstechnica.com/space/2026/0...

I just recently submitted this related abstract to NASA's Mission Software for the Science Mission Directorate Workshop 2026. (science.data.nasa.gov/features-eve...)

jonny (leachate composite form)@jo.nny.rip · last wk.

If you want to "just do biology and never worry about programming" the answer looks like adequate funding for research software engineering so we can make validated, shared tools that reduce the need for new code to none. Not everyone generating analysis on demand that is impossible to validate.

Schweickhart and Bolden and the rest: Thank you. www.theguardian.com/science/2026...

‘Our country is the mission’: more than 100 ex-Nasa astronauts launch new voting campaign

Astronauts for America liken effort to instill trust and faith back in constitution’s principles and democratic institutions to ‘making a course correction’

theguardian.com

One of the biggest and most disruptive innovations of "new space" is that they no longer meticulously define mission success criteria well in advance, with all stakeholders involved, in unambiguous language and verifiable, quantified benchmarks. Now it's vibes. Was that a success? 🤷 Who's asking?

One of the biggest and most disruptive innovations of "new space" is that they no longer meticulously define mission success criteria well in advance, with all stakeholders involved, in unambiguous language and verifiable, quantified benchmarks. Now it's vibes. Was that a success? 🤷 Who's asking?

Went well enough. It would be more impressive if this was Flight 2, not 13. Still losing tiles, there's damage to the airframe, and Super Heavy was lost again. On the same flight path as the others. At some point, this thing has to actually work.

Big Dumb Booster@buccaneerboosters.bsky.social · 3w ago

To reiterate: it's abundantly clear the Starship program is a colossal systems engineering failure. Despite cries of "iteration", we are at the thirteenth flight and third version of this vehicle and it still lacks basic functionality.

As a fun fact. Starship has now launched as many times as the Saturn V. Starship has yet to make orbit.

Trump Administration Rules Make Research Slower, Costlier and Harder to Share scienceblogs.com/sb-admin/202...

Trump Administration Rules Make Research Slower, Costlier and Harder to Share | ScienceBlogs

Most Americans don’t read the rules that govern the federal grants funding science. Those rules, called the Uniform Guidance, shape how cancer research is funded, graduate students are trained, clinic...

scienceblogs.com

booster fails to relight engines for landing burn, and explodes on impact with the Gulf of Mexico

As *tech*, AI isn't exceptional. It's not exceptionally wicked. It's not exceptionally good. Take away the accompanying, galactic-scale stock-swindle, and we'd call AI's apps "plug-ins" and we'd use them and abuse them in the same way that we've used every other technology: www.normaltech.ai 1/

NASA programs feel effects of workforce reductions More than two dozen major NASA projects say they have been affected by the loss of a fifth of NASA’s civil servant workforce in the last year....

NASA programs feel effects of workforce reductions

More than two dozen major NASA projects say they have been affected by the loss of a fifth of NASA’s civil servant workforce in the last year.

spacenews.com

By training and culture, scientists tend to assume that others are acting generally in good faith, not willfully lying, and share an interest in pursuit of truth for the good of humanity. This is easily exploited by hostile and bad faith actors. Sometimes the correct response is "Fuck you, liar."

Derek Lowe@dereklowe.bsky.social · 2w ago

Thoughts on the various Trump administration proposals to revamp US science funding - and what they all have in common:

I'm sure it's a complete coincidence that this also puts the launch attempt after the market closes for the weekend.

SpaceX is postponing Starship IFT-13 one day. Now July 24. Says on X they want clear imagery from the ground of Starship's heat shield "as it flies at a higher dynamic pressure during ascent" and weather today isn't good for that. x.com/SpaceX/statu...

The Economist interview of Elon Musk is astounding in all the ways you'd expect. Musk is unapologetic, combative and unwilling to accept or consider facts stated by the Editor in Chief who interviews him including about her own country. www.economist.com/leaders/2026...

Should you be afraid of Elon Musk?

Artificial intelligence is charging ahead. Not even its creators know how to keep up

economist.com

A devastating but lesser known part of federal science attacks is on the federal science advisory system. In a new post, I discuss the state of play on federal advisory committees and how the scientific community is pushing back through independent science efforts. blog.ucs.org/gretchen-gol...

"How did the bot escape its vibe coded sandboxing? [...] Well, they vibed the sandbox, you tell me. [...] If the story told by these press releases is 100% true, then OpenAI hacking HuggingFace is when incompetence crashed into incompetence and they turned it into a mutual marketing stunt."

OpenAI hacks HuggingFace with an AI — allegedly

On 16 July, the AI model repository HuggingFace posted a security incident report. Well, I say “incident report” — it’s written like an AI doomsday press release: [GitHub] Earlier this week, we det…

pivot-to-ai.com