State of Digital Discourse: What Google Analysis Reveals About Forum Toxicity

[LinkedMusicians]

Marketers have been using sentiment analysis to better understand consumer data since the 1990s. Google has been refining it — using it for analyzing what becomes search results — for more than a decade. Google’s Natural Language Processing (NLP) quantifies emotional tone and attitudes from text. So I provided Google’s Gemini AI tool with uniform criteria to analyze and measure toxicity and the politicization of KVR, GearSpace, VI-Control, Cakewalk Discuss Forums, and LinkedMusicians, and put the results side-by-side in a chart. I designed LinkedMusicians based on being the antithesis of the toxicity, bias, trolling, sniping, culture wars, bigotry, and developer- and advertiser- focused censorship of those forums, and one year after our public launch, I wanted to see how Google’s AI finds we compare to the legacy forums. It’s one thing to build ethical and non-toxic policies for a community; effectively implementing those policies is a completely separate challenge. I believe that we’re on track. But I wanted to know what Google’s analysis found.

In the blog post, I share the unedited results of the analysis. Please check out the blog post and share your thoughts below. What  do you think? Does being in an environment where you won’t be trolled or censored for a post sharing a negative opinion of a developer site owner, advertising, or affiliate partner matter to you?  Do you find that toxic environments stifle open and honest conversations?

[See the full post at: The State of Digital Discourse: What Google’s AI Analysis Reveals About Popular Music Producer Forums]

CURATED BEST FREE LISTS: Kontakt Libraries, Sample Libraries, Virtual Synths, Effects
POWERSEARCH TOOLS: Synth Presets, Loops & Sound FX, Gear Reviews 
FREE TOOLS: Mixing Problem Solver, Songwriter's Toolbox, QuickGuides, Scale, Mode & Chord Finder
READ/WATCH: Mag Articles / Podcasts, Audio Production & Songwriter's Dictionaries

2

[LinkedMusicians]

On a lighter note, I’ve trained Gemini not to give the kind of praiseful comments popular Large Language Models (LLMs) are programmed to give users. I want facts and analysis, not programmed compliments. Subsequently, the results are often brutally frank. For example, I cut and pasted in my first draft of the blog post linked above  and Gemini responded that it was excellent, but that it read like a “white paper.” The response made me smile. I realized that it was right. I tend to write out facts in a very analytical manner on a first draft and later come back to inject personality and humor into content by the second or third round. Consequently, I revised the post to be a lighter, more enjoyable read, and made the headline less dry and more newsworthy.

CURATED BEST FREE LISTS: Kontakt Libraries, Sample Libraries, Virtual Synths, Effects
POWERSEARCH TOOLS: Synth Presets, Loops & Sound FX, Gear Reviews 
FREE TOOLS: Mixing Problem Solver, Songwriter's Toolbox, QuickGuides, Scale, Mode & Chord Finder
READ/WATCH: Mag Articles / Podcasts, Audio Production & Songwriter's Dictionaries

1

[Author]

Posts