MSD / JOURNAL / RESEARCH EDITION

Fortschritt.
Mit Belegen.

All topics, subpages and portals in one place. Discover the MSD world via the illustrated introductions and the complete table of contents directly below.

RECHERCHESTAND 12.11.2026 · FAKTEN + EINORDNUNG + SZENARIEN
FIELD NOTES / OPEN QUESTIONS / REAL SOURCES3.332

Research and literature notes

ERFOLGE. GEGENBEFUNDE. OFFENE FRAGEN.
LEARNTESTFAILREBUILD

MSD / DAS INHALTSVERZEICHNIS

Alles auf einen Blick.
Alles miteinander verbunden.

Choose one of the most important areas. Below you will find the complete list of all recorded public entries - directly on this page, without having to open the main menu first.

The complete directory

261 public entries and subpages; therein 65 collections. Additionally 792 subject filter. The count refers to entries and filters, not the number of available offers or own research projects.

MSD & Journal 42 entries · expand
AI, robotics & future 7 entries · expand
MSD Direct & Marktplatz 27 entries · expand
Portale & Kollektionen 137 entries · expand
Service, Konto & Vertrauen 34 entries · expand
Die MSD-Familie 9 entries · expand
Ausblick & Entwicklung 5 entries · expand
Alle Fachgebiete 792 entries · expand

MSD / JOURNAL / DARUM GEHT ES

Die gesamte MSD-Welt. Direkt zum richtigen Inhalt.

01

Orientation across all areas

The main topics are visible directly below as illustrated introductions. The complete directory also bundles all 261 recorded public entries and subpages as well as 792 subject filters. Offers, planned areas and editorial content retain their respective labels.

02

Classify research in an understandable way

26 research notes explain specific findings, limitations and possible consequences for companies. This includes positive productivity findings as well as counterexamples and failed expectations. Every classification leads to the source; external research is not presented as a separate MSD test.

03

Eine Bibliothek zum Weiterarbeiten

3,306 clear references link publications on AI, robotics, work and technical systems via their DOI. You can search the research archive by topic and term. The metadata is a working library; they do not replace checking the respective full text.

01 / DIE REDAKTIONELLE SPUR

Keine Erfolgsgeschichte ohne Gegenprobe.

26 classified research notes and 3,306 DOI linked references. External findings, editorial derivations and planned MSD experiments remain visibly separate.

Wirtschaft & Arbeit / Befund / 2023

KI im Support: messbarer Nutzen, ungleich verteilt ↗

5,179 Support staff form the basis of the NBER working paper. With assistance, the number of issues resolved per hour increased by an average of 14%. New and less experienced forces particularly benefited.

Brynjolfsson, Li & Raymond · NBER 31161
AI & Agents / Contrary finding / 2025

When AI makes work take longer: METR's contrary finding ↗

In a randomised trial with 16 experienced developers and 246 tasks, early-2025 AI tools increased completion time by 19%. Participants perceived their benefit much more favourably.

METR · Randomised trial
KI & Agenten / Methodik / 2026

Neue Werkzeuge, neue Auswahlverzerrung ↗

METR describes a follow-up experiment in February 2026. Evidence of faster work is difficult to interpret due to selection of participants and tasks as well as problems with time measurement.

METR · Methodenupdate 24.02.2026

02 / DAS DURCHSUCHBARE ARCHIV

Tausende Spuren. Ein Einstieg.

The archive can also be read without JavaScript via the linked pages.

Wirtschaft & Arbeit / Befund / 2023

KI im Support: messbarer Nutzen, ungleich verteilt ↗

5,179 Support staff form the basis of the NBER working paper. With assistance, the number of issues resolved per hour increased by an average of 14%. New and less experienced forces particularly benefited.

Brynjolfsson, Li & Raymond · NBER 31161
AI & Agents / Contrary finding / 2025

When AI makes work take longer: METR's contrary finding ↗

In a randomised trial with 16 experienced developers and 246 tasks, early-2025 AI tools increased completion time by 19%. Participants perceived their benefit much more favourably.

METR · Randomised trial
KI & Agenten / Methodik / 2026

Neue Werkzeuge, neue Auswahlverzerrung ↗

METR describes a follow-up experiment in February 2026. Evidence of faster work is difficult to interpret due to selection of participants and tasks as well as problems with time measurement.

METR · Methodenupdate 24.02.2026
Economics & Work / Context / 2025

Exposure is not a layoff forecast ↗

The refined ILO index examines which occupational activities could be exposed to generative AI. It assesses technical exposure, not actual decisions to cut jobs.

ILO · Working Paper 140
Wirtschaft & Arbeit / Mechanismus / 2018

The J-curve: rebuild first, harvest later ↗

Brynjolfsson, Rock, and Syverson explain how complementary investments in organization and knowledge can delay the measured benefits of an enabling technology.

NBER · Productivity J-Curve
Wirtschaft & Arbeit / Gegenbefund / 2017

Industrial automation can displace employment ↗

Acemoglu and Restrepo examine historical US labor markets with varying exposure to industrial robots and find negative local employment and wage effects.

Acemoglu & Restrepo · NBER 23285
Wirtschaft & Arbeit / Modell / 2024

Why a strong bot doesn't prove an economic boom ↗

Acemoglu's task-based model connects affected activity shares with savings and overall economic productivity. The May version 2024 estimates at most 0.66% additional TFP over ten years.

Acemoglu · Modellfassung Mai 2024
Systeme & Infrastruktur / Infrastruktur / 2026

The physical limit of digital expansion ↗

The IEA describes 2026 increasing power demands from data centers and bottlenecks in networks, chips and financing. Efficiency gains per task can be outweighed by increased usage.

IEA · Key Questions on Energy and AI 2026
Wirtschaft & Arbeit / Grundlage / 2024

Money is an entitlement – not a robot battery ↗

The ECB differentiates between central bank money and commercial bank money. Lending creates a deposit; Redemption makes this book money disappear again.

EZB · What is money?, aktualisiert 2024

{link("/company/de/journal/archiv/","Alle Literaturseiten ohne JavaScript")}

Wie dieses Archiv entsteht

The editorial notes are based on the primary sources linked. The additional bibliography uses publicly deposited Crossref metadata: title, authorship, year of publication, place of publication and DOI. Cleaned for DOI and normalized title; Title comparison for four subject areas, period 1990–12.11.2026. No abstracts copied. No success rates, failed attempts or fabricated content from a title. Not every reference is editorially read, independently reviewed or peer-reviewed; Conference papers and book chapters may also be included. Publication dates may vary depending on the publisher. The index is a comprehensible working library, not a systematic review.

Your own laboratory tests require a real test protocol. So far, no own test successes or failures have been claimed here.

{link("https://www.crossref.org/documentation/retrieve-metadata/rest-api/","Crossref-Metadaten and API")}

03 / THE NEXT TRY

A clean miss is a result.

01

Aufbau festhalten

Document task, version, material, environment and baseline before start.

02

Count interventions

Not just successful videos: also measure interruptions, repetitions, rework and operating time.

03

Erkenntnis weitergeben

Publish results and limitations. Only real, documented tests are referred to as MSD tests.