Will Depue stated that Claude agents identified dwarf-planet candidates in public NASA archives over a few weeks of automated scanning. Alongside his collaborators, he is now funding further amateur astronomy initiatives with $100,000 in grants through his nonprofit, Macro Grants.
Searching Public Archives With Claude Agents
Government-funded telescopes operated by American, European, and Japanese agencies image the night sky constantly and publish those gigabytes of data online. Most surveys focus on specific targets, leaving vast amounts of archival data unexamined for wandering objects. Depue, alongside collaborators Luke Farritor and Josh Wall, utilized Claude and Codex agents to comb through these public astronomy repositories over the course of a few weeks.
“There’s just tons of this data out there and a lot of it hasn’t been totally deeply finely combed through because people just don’t have the time to look for every possible object that appears in any of these data sets,” Depue said on the MTS podcast, according to BigGo Finance.
The search originally began as an attempt to find Planet 9, the hypothetical ninth planet of the solar system, before drifting into other discoveries. The project yielded asteroid and dwarf-planet candidates, alongside a series of unclassified items. However, the exact count of discoveries, verification pipelines, and specific object designations remain unreleased.
Human Attention Constraints Cause Data Gaps
When questioned about why NASA failed to locate these objects using identical cloud access, Depue pointed toward human attention constraints rather than data availability. Astronomical institutions direct their limited talent toward higher-priority targets, leaving the long tail of small solar-system objects largely ignored.
“OpenAI only published the top 400 most important results they got,” Depue said.
“There’s a massive number of unimportant results they didn’t publish or that people independently are discovering.”
This institutional blind spot forms the foundational thesis behind Macro Grants, a nonprofit aimed at funding whimsical research projects.
The organization offers grants ranging from $500 to $25,000, promising decisions within two to three days. The newly announced $100,000 amateur astronomy grant is the first public instance of a larger fundraising push designed to support similar global work.
Matching Capital to Compute and Inputs
Depue frames these projects as functions of specific inputs, such as compute, dollars, and tokens. “A lot of the hard work is coordinating with governments and getting the plane in the air.”
The nonprofit’s review board includes projects like Mecan’s LIDAR archaeology flights over Central American jungles, Luke Farritor’s decipherment of Herculaneum scrolls, and Riley Walls’ San Francisco scavenger hunt. Depue argues that donors feel more engaged when financial contributions translate instantly into immediate discoveries rather than traditional, slow-moving academic pipelines.
Automated Systems May Outperform Human Skills
The conversation during the podcast eventually shifted toward long-term artificial intelligence trajectories and potential civilizational risks over the next thirty to forty years. Depue compared future human skill sets to chess player Magnus Carlsen facing the Stockfish chess engine, predicting that functional human skills will be entirely outperformed by automated systems.
“We’re headed towards a world in which I am like you know would be equivalent to what Magnus Carlson is to Stockfish 14,” Depue said.
He described this shift through a natural selection lens, arguing that any system capable of self-replication and self-modification inherently triggers evolutionary pressure. Machines possess clear structural advantages over biological computers, including rapid copying, energy efficiency, and arbitrary energy expenditure. While acknowledging that individual humans will likely persist, Depue argued that humans will not remain the dominant species over extended cosmic timescales.
Funding the Replacement of Human Researchers
An underlying irony defines the entire project: the automated capability Depue is currently financing to out-search human researchers in public datasets is the exact technology he predicts will render human researchers obsolete. “People should take everything I say with a grain of salt. I tweet everything like in 30 seconds,” Depue said while discounting his own long-term forecasts.
Market valuations currently reflect high expectations for AI-driven science, with Anthropic targeting a $1.4 trillion listing and Nvidia reporting guided revenue of $108 billion with 74% gross margins. Whether Depue’s dwarf-planet candidates survive rigorous astronomical verification remains unconfirmed.
