SYNTHESIS NOTE
TopicsNovel Architecturesthis note

Can AI systems improve themselves through trial and error?

Explores whether replacing formal proof requirements with empirical benchmark testing enables AI systems to successfully modify and improve their own code iteratively, and what mechanisms prevent compounding failures.

Synthesis note · 2026-02-23 · sourced from Novel Architectures

The original Gödel Machine proposed self-improving AI via provably beneficial self-modifications. In practice, formally proving the impact of most self-modifications is impossible. The Darwin Gödel Machine (DGM) replaces formal proofs with empirical validation: try modifications, test them on benchmarks, keep what works. This mirrors biological evolution — mutations are not verified in advance but produced, trialed, and selected.

DGM alternates between self-modification and evaluation phases. During self-modification, agents from the archive generate modified versions of themselves — rewriting their own code. During evaluation, each modified agent is tested on coding benchmarks. The key assumption: improvement on coding benchmarks indicates better coding capabilities, which in turn indicates better ability to self-modify. This creates a meta-competence loop: better coding → better self-modification → better coding.

Results: SWE-bench from 20.0% to 50.0%, Polyglot from 14.2% to 30.7%.

The evolutionary archive is critical. Inspired by open-endedness research, DGM maintains a growing library of all generated agent variants — including suboptimal but interesting ones. These serve as stepping stones for future generations, enabling diverse exploration paths. The system doesn't just optimize for immediate performance; it accumulates diverse capabilities that may enable future breakthroughs. This is fundamentally different from single-trajectory self-improvement.

Concrete improvements discovered include better code editing tools, long-context window management, and peer-review mechanisms — capabilities the original agent lacked that emerged through the self-improvement process.

The Python-based implementation makes the self-modification space Turing-complete in principle. The current version modifies agent design (tools, workflows) with frozen foundation models. Full self-improvement — rewriting training scripts, training new foundation models — is left as future work.

This directly addresses What limits how much models can improve themselves?: DGM circumvents the formal proof requirement by using empirical validation, but inherits a different limitation — improvement is bounded by what the benchmark can measure. The archive approach partially addresses How quickly do errors compound during model self-training? by maintaining diverse populations rather than following single improvement trajectories.

Inquiring lines that read this note 100

This note is a source for these research framings, grouped by the broader line of inquiry each explores. Scan the bold lines of inquiry; follow any specific question forward.

How do we evaluate AI systems when user perception misleads actual performance? Why does verification consistently lag behind AI generation? Can AI-generated outputs constitute genuine knowledge or valid claims? Does AI fluency substitute for verifiable accuracy in human judgment? How can humans calibrate appropriate trust in AI systems? How should models express uncertainty rather than forced confident answers? How can AI systems learn from failures without cascading errors? Does self-reflection enable models to reliably correct their errors? How does objective evolution guide discovery better than fixed planning? When should tasks involve human-AI partnership versus full automation? How do multi-agent systems achieve genuine cooperation and reasoning? Can inference-time compute substitute for scaling up model parameters? How effectively do deterministic tools improve language model reasoning on formal tasks? Why does self-revision increase model confidence while degrading accuracy? Which computational strategies best support reasoning in language models? Why do self-improving systems struggle without clear external performance metrics? Does parallel reasoning outperform sequential thinking under fixed compute budgets? How does example difficulty affect learning efficiency in language models? Why do agents confidently report success despite actually failing tasks? What capability tradeoffs emerge when scaling model reasoning abilities? Does externalizing cognitive work and state improve agent reliability? When does optimizing for quality undermine the value of diversity? Can single-axis benchmarks accurately predict agent deployment success? How do self-generated feedback mechanisms enable effective model learning? How do evaluation mechanisms prevent error accumulation in autonomous research systems? How can AI agents autonomously learn and transfer skills across tasks? How should systems govern persistent agent-generated code in shared infrastructure? How should human oversight be integrated with autonomous AI systems? How should iterative research systems allocate reasoning per search step? Does fine-tuning modify underlying model capabilities or only behavioral outputs? Why do benchmark improvements fail to reflect actual reasoning quality? Do harness improvements transfer across model scales or memorize shortcuts? What are the consequences of models training on synthetic data? How do training data properties shape reasoning capability development? Can language model RL training avoid reward hacking and misalignment? How can identical external performance mask different internal representations?

Related concepts in this collection 6

This note in its neighbourhood — explore the map, then jump to a related concept in the list below.

Concept map
18 direct connections · 133 in 2-hop network ·medium cluster Open in graph ↗

Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph

your link semantically near linked from elsewhere

Related papers in this collection 8

Papers most semantically related to this note, ranked by cosine similarity in the embedding space.

Original note title

darwin godel machine achieves open-ended self-improvement by replacing formal proofs with empirical validation and evolutionary archives