Recently, Altbot has been targeted by DDoS attacks. While the motive is unclear, it seems tied to a misguided belief that the bot’s alt-text generation is harmful to the environment. Let me set the record straight.
Altbot uses Gemini-1.5-flash, a specifically low-power AI model. Processing a single alt-text request consumes around 0.0005 kWh, meaning that in the 4 months since I created Altbot—and tens of thousands of alt-text entries generated—it has consumed roughly 10 kWh in total. That’s about the same energy as running a single LED light bulb nonstop for 41 days or driving an electric car just 40 miles.
These DDoS attacks, however, have already consumed more electricity and computing power than Altbot has across its entire existence... in a matter of hours. This means the attacks have caused more environmental harm than the very thing they seem to be protesting.
Altbot exists to improve accessibility on Mastodon, not to harm the environment. Accessibility and sustainability can coexist, and I’m committed to keeping Altbot energy-efficient and purposeful.
If there are concerns, I encourage constructive dialogue—not destructive actions that undermine both inclusivity and sustainability.
Thank you to everyone supporting Altbot’s mission of making the Fediverse more accessible for all.
@micr0 damn we really do have to tap the sign don't we
@ari @micr0 @Makura To give a concrete example, unless it was trained from data where all use of gendered words was intentionally stripped, image labeling models will necessarily misgender people in the images they describe based on biases regarding gender roles or gender presentation.
They'll also do the opposite, e.g. mislabeling a doctor as a nurse because of feminine appearance. This direction of error is likely to have racist aspects too, and can happen from sampling bias even if words mentioning gender and race were scrubbed from training data.
All of this is a very legitimate reason to hate "AI" even in "benign" and local contexts.
@dalias @ari @micr0 @Makura also, quoting how much electricity is used in use is misleading.
How much water and power was used by the creation of Gemini? This needs accounting for too.
And while building dependency on billionaires' products, you need to consider the harm those are going to be doing thanks to projects normalising and (indirectly) funding them. eg google have recently abandoned CO2 reduction targets bc AI.