Microsoft has released an Excel update to address an auto-formatting feature that was scrambling gene names, forcing scientists to alter conventions to avoid data corruption. The change gives users more control over automatically converting text strings.
Update Adds Option to Disable Problematic Conversions
On Thursday, Microsoft detailed a new Excel option that lets users toggle the auto-formatting of text strings into dates. This change directly targets the gene naming issue.
Read Also: Elon Musk Billion Offer Wikipedia Name Change
Scientists had used alphanumeric shorthand like “SEPT1” for genes. But Excel converted these into dates like “September 1st”, disrupting gene data. The update lets users disable such automatic conversions.
Auto-Formatting Can Corrupt Scientific Data
Excel auto-formatting aims to accelerate entering common data like dates. But it can wreak havoc when shorthand text unintentionally triggers conversions.
A 2016 study found gene names were getting converted to dates, undermining published research. To avoid this, scientists started using symbols or changing conventions – a major hassle.
New Setting Builds on Previous Customization Options
Last year, Microsoft gave users more control over auto-formatting with options to be warned before conversions or load files without them.
Read More: Adobe MAX 2023 New Firefly AI Updates Creative Cloud
The new toggle to entirely disable text-to-date conversions enhances user control further. But Microsoft notes limitations, like stored data losing calculation capabilities.
Workaround Caused Headaches for Scientists
The gene naming issue made global headlines in 2020 when scientists revealed they altered conventions to avoid automatic formatting.
This clumsy workaround caused major disruptions. Researchers had to overhaul documentation and workflows due to Excel’s unwanted “help”.
Bug Highlighted Over-Eagerness of AI Features
The incident highlighted pitfalls of AI features aggressively transforming content without asking. Excel’s logic short-circuited when applied universally.
It demonstrated the need for nuanced settings that consider outliers like shorthand gene names. Users require visibility into how AI alters data.
Scientists Welcome Fix But Want More Transparency
Researchers have welcomed the update, but emphasize a broader need for transparency into how software handles data.
Give users visibility when automatic changes occur and involve domain experts when training AI to avoid unintended harm to specialty use cases.
Overall, this fix gives scientists needed relief from disruptive auto-formatting. But it’s just one step toward AI systems collaborating more thoughtfully with human expertise.