AI is here, but it won’t save us.
Wrestling with complicated feelings surrounding AI? Us, too.
Much of the conversation around AI is a binary of yes always or no never. Like many things, the binary is insufficient. We are all using AI whether or not we like it. It shows up in our search engines, and companies sell (or already sold) our data to train AI language models without our consent.
AI is also a limited tool. It’s incentivized to provide an answer even in the absence of accurate information. It will spit out wrong answers over simply saying it doesn’t know. A study by Columbia Journalism Review showed AI-chat bots were confidently wrong 60% of the time.
AI is here, but it won’t save us. We believe regulation is necessary to curb organizations selling our data, address the inherent bias that harms marginalized communities, and force AI companies to provide transparency and accountability on energy usage and its negative environmental impacts. There’s a fantastic article on the complicated question of AI’s environmental impact from MIT Technology Review.
Do we still feel weird about it? Yes. Will we change our minds every five minutes on whether the ends justify the means? Also yes.
Why are we using AI?
We are using AI to take your words and transform them into something unique, something that will get past the algorithm gods and into people's newsfeeds.
We insist (as much as we can) that you proofread the AI text for errors.
This is an opt-in option. If you keep it off, the tool will use the words you entered as-is, and will pull a unique image for your background.
This will help scoot your posts under the radar. As per usual, much of social media is a constantly evolving mystery, and we do our best to adapt to these changes.
Because this is our country, too.
Which AI do we use? Is it ethical?
We work with Mistral AI at the moment. This field is changing all the time, and we will pick the best option with the least impact possible.
However, no AI can be considered 100% ethical. We chose Mistral AI because it’s based in the EU, which has more oversight than the United States.
EU countries must comply with stricter environmental controls and privacy standards. In short, there’s more regulation, which we support.
They are also the only company that's
released a comprehensive report on the environmental impact of their AI models and their lifecycle
.
On their website, they state,
“without more transparency, it will be impossible for public institutions, enterprises and even users to compare models, take informed purchasing decisions, fill enterprises' extra-financial obligations or reduce the impacts associated with their use of AI.”
We use Mistral’s small language model because smaller models use less energy and resources.