Good - Bad AI? c/o FT Magazine et al.

My source: Tim Harford, How can we tell good AI from bad? FT Magazine. 21 March 2026. #1168: pp.9-10.

    'Two new working papers address the tricky issue of verifying quality. In "Some Simple Economics of AGI", Christian Catalini, xiang Hui and Jane Wu (assisted, sometimes gratingly, by generative AI) propose the inevitable 2x2 matrix in which economic activity can be easy to automate, easy to verify, both or neither. Automatable, verifiable output is the stuff that computers do for us. The non-automatable stuff remains reassuringly artisanal.
    The difficult quadrant is where tasks seem easy to complete but are hard to check. Catalini, Hui and Wu call this the "runaway risk zone". It is not a reassuring label and it is not meant to be.
...
    In the second paper, "A model of Artificial Jagged Intelligence", Joshua Gans offers an analogy in which asking AI to perform task is like trying to cross a river over a network of planks supported by occasional pylons. The jagged frontier is represented by the fact that some planks are long and wobbly, while others are short and sturdy. Problem one: even if the planks are typically sturdy, the wobbly planks will require most of your time and attention. Problem two: if you can't predict in advance which planks will let you down, you may quite sensibly prefer to eschew the AI entirely and row yourself across the old-fashioned way.'

Catalini, C., Hui, X., & Wu, J. (2026). Some Simple Economics of AGI. arXiv preprint arXiv:2602.20946.

Joshua S. Gans, 2026. "A Model of Artificial Jagged Intelligence," NBER Working Papers 34712, National Bureau of Economic Research, Inc.

Previously: 'safety' : 'ai' : 'domain' : 'matrix'

Workshop on Research Objects (RO2018)

2018-10-29 at IEEE eScience 2018, Amsterdam, The Netherlands
http://www.researchobject.org/ro2018/

## Timeline
...

## Research Objects

Scholarly Communication has evolved significantly in recent years, with an
increasing focus on Open Research, FAIR data sharing and community-developed
open source methods. A question remains on how to publish, archive and explore
digital research outputs.

A number of initiatives have begun to explore how to package and describe
research outputs, data, methods, workflows, provenance and structured metadata,
reusing existing Web standards and formats.

Such efforts aim to address the challenges of structuring multi-part research
outcomes with their context, handling distributed and living content and
porting and safely exchange what we collectively can call “Research Objects”
between platforms and between researchers.

## Call for Papers

In the workshop RO2018 we will explore recent advancements in Research Objects
and publishing of research data with peer-reviewed presentations, invited
talks, short demos, lightning talks and break-out sessions to further build
relationships across scientific domains and RO practitioners.

RO2018 welcomes submissions of academic abstracts (1-2 pages), data/software
articles (3-6 pages) and short research articles (4-8 pages) on cross-cutting
case studies or specific research on topics including, but not limited to:

  FAIR metrics; platforms, infrastructure and tools; lifecycles; access control
  and secure exchange; examples of exploitation and application; executable
  containers; metadata, packaging and formats; credit, attribution and peer
  review; dealing with scale and distribution; driving adoption within current
  scholarly communications and alignments with community efforts; and
  domain-specific and cross-domain Research Objects.

## Submitting

Submitted abstracts and articles can be in a range of open formats (e.g. HTML,
ePub) and are particularly encouraged to be submitted in a FAIR research data
packing format.

Accepted abstracts and articles will be included in the IEEE eScience 2018
proceedings. Submitted preprints will, upon acceptance, be made available as
Green Open Access on the RO2018 website with DOI links to the Zenodo record and
(eventually) the published IEEE proceeding article.

It is a requirement that at least one author of each accepted submission
attends the RO2018 workshop and at the IEEE eScience 2018 conference, where
registration fees applies.

Further details on submitting ...
...

Stian Soiland-Reyes (my source: by email)
The University of Manchester
http://www.esciencelab.org.uk/
http://orcid.org/0000-0001-9842-9718