FY 2026-27

AI Research Diary

Dated technical records of specific research questions, background searches, hypotheses, experiments, results and conclusions.

ENTRY STRUCTURE core activity
01Hypothesisstate a testable prediction
02Experimentrecord method and controls
03Evaluationcompare results with criteria
04Conclusionrecord what was established
background researchrecorded conclusion
5research entries
3research projects
0currently running
Gitcommit history retained

Latest entries

Recent research

View all entries →
RD-2026-008 concluded

Does the same agent loop produce the same investigation?

Three non-frontier coding models reviewed the same pre-fix software through a common agent loop. Their execution histories exposed different evidence choices, missed boundaries and unsupported conclusions, showing why tool availability alone does not establish investigative reliability.

19 Sep to 20 Sep 2026 2 days 12 research hours · estimated Core
#ai#coding-agents#non-frontier-models#verification
RD-2026-007 concluded

When the compiler writes itself, what is the source?

We replaced an implementation-shaped natural-language corpus with five small Program files, rebuilt ProgSync without its former source, and tested the result against an independent 55-case oracle. The build worked, but one clear synthesis error showed why verified generated code cannot be disposable.

23 Jul to 23 Jul 2026 1 days 15 research hours Core
#ai#natural-language#program#progsync
RD-2026-004 concluded

Can we turn source code into natural language?

We tested whether readable Markdown could preserve enough of a program for isolated AI translators to reconstruct working software. A 21-module trial succeeded; self-hosting exposed deeper contract failures and left the central claim unproven.

21 Jul to 24 Jul 2026 4 days 32 research hours · estimated Core
#ai#natural-language#program#progsync

Record-keeping approach

How entries are maintained

The diary records when decisions were made, how experiments were run and which files support the reported results.

01

Record plans first

Background searches, hypotheses and evaluation criteria are committed before experimental work starts.

02

Keep complete results

Entries include unsuccessful runs, unexpected behaviour and results that do not support the hypothesis.

03

Use stable references

Every entry has an identifier, dates, named investigators and a link to its project and core activity.