# Paper for discussion: Making AI reliable

**URL:** <http://discourse.iapct.org/t/paper-for-discussion-making-ai-reliable/16271>\
**Category:** AI\
**Created:** [October 3, 2026, 12:39am UTC](http://discourse.iapct.org/t/paper-for-discussion-making-ai-reliable/16271 "2026-10-03T00:39:59Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![bnhpct](http://discourse.iapct.org/user_avatar/discourse.iapct.org/bnhpct/32/34_2.png) [@bnhpct](http://discourse.iapct.org/u/bnhpct)\
**Post date:** [October 3, 2026, 12:39am UTC](http://discourse.iapct.org/t/paper-for-discussion-making-ai-reliable/16271/1 "2026-10-03T00:39:59Z")

</div>

I have just published this for discussion at the 2026 conference in York:

Nevin, B., & Claude (Anthropic). (2026). Making AI reliable. Zenodo. [Making AI reliable | Zenodo](https://doi.org/10.5281/zenodo.23111606)

---

<div class="post-metadata">

**Author:** ![bnhpct](http://discourse.iapct.org/user_avatar/discourse.iapct.org/bnhpct/32/34_2.png) [@bnhpct](http://discourse.iapct.org/u/bnhpct)\
**Post date:** [October 30, 2026, 9:24pm UTC](http://discourse.iapct.org/t/paper-for-discussion-making-ai-reliable/16271/2 "2026-10-30T21:24:58Z")

</div>

A revision of the paper. For version 1, Zenodo currently reports 8 downloads and 41 views.

Nevin, B., & Claude (Anthropic). (2026). Making AI reliable. Zenodo. 36th Annual Conference of the Association for Perceptual Control Theory (IAPCT), York, UK. [Making AI reliable | Zenodo](https://doi.org/10.5281/zenodo.23809157)

I think my parts of it are now clearer and more readable, and some official requirements are now properly discharged, notably credit to the NSF in the requisite form so I can now cite it in my final NSF reports.

I intended this for distribution at our IAPCT conference (whence the credit to IAPCT in the citation), but the paper did not start taking shape until after the schedule of presentations was full and closed.
