One Red Flag referenced here that I’ve seen in many districts is trading sponsorship for access. A company will sponsor an event put on by a public orgs sister non-profit. The non-profit gets paid. Money goes to public workers via the non-profit. And the sponsoring company gets a contract.
One public school system even set budgets of AI contracts based on what that they could juice from sponsorships.
This article largely came out of frustration with my peers investing lots of energy in developing LLM accessibility review. I spent a year doing it myself. Now I'm done. (Props to those we reference in the article.) My focus is now on quicker scalability for rules-based accessibility tests.
LLM-generated descriptions miss lots of context. For instance, depending on the site and content, we might mention people's races or fashion. Other times we don't.
everyone should be taking this stuff seriously. EU's accessibility act is set to be implemented in 2025 and the US congress also has a bill in the works that mandates greater accessibility across all governments.
We're pre-revenue. Our salaries would be covered by a few interested investors. Current thinking: Unless we get a strategic benefit from taking money from the investor, we shouldn't.
I'm really just curious if some data exists that shows startups that paid salaries survive longer than startups that don't.
Ya. Totally. The good manager is my Plan B if I don't get an offer I want. We have loyal clients that most companies are hot for + a great workflow, so hopefully that translates into more than just a job..
Super solid. We have a workflow that allows devs or designers with under 3 yrs of experience to develop sites that usually require pros with > 5 yrs. That means lower overhead of us and stronger profit.
In my experience, the best policy is just to ask engineers where they are spinning their wheels. Engineers know best where to automate. Relying on top down automation decisions inspires ineffective automation.
A few months ago, I had to build themes for 20+ sites. The sites varied in features (some had slideshows, some had blogs, ..) I wanted to build the projects from boilerplate code that could adapt to the various features as needed, but no tool existed. I pulled in my brother, @ebertucc, and we created Shaper. Now I spend 1-2 minutes instead of 2-3 hours setting up code projects.
The app was built in a few months using Node, Vue and Electron. We're launching this Alpha version to see if Shaper is something we should develop beyond our immediate needs (a few lawyer friends are clamoring for Shaper to segment and generate contract documents).
Blake is indeed great summer reading - a retreat from the rational mind. In that vein, my two summer picks are A.R. Ammons' "Still" and the fragments of Friedrich Holderlin translated by Michael Hoffman.
One public school system even set budgets of AI contracts based on what that they could juice from sponsorships.
The solution? Pay public employees more?