Parents—have you seen the reports linking the hepatitis B vaccine to serious autoimmune conditions like lupus, rheumatoid arthritis, and type 1 diabetes? While these cases are rare, staying informed is essential when making health decisions for your child. Here’s a resource that breaks it down: #vaccines#

Did you know that the hepatitis B vaccine has been linked to reported cases of severe autoimmune diseases, including systemic lupus erythematosus, rheumatoid arthritis, and type 1 diabetes? Staying informed is key to making the best decisions for your child’s health. Here's a good resource: #vaccines#

Major congrats to the legend @SugarRayLeonard and his foundation on the big night! Thank you for all the incredible work that you do funding research, care and raising awareness for pediatric type 1 & 2 diabetes.

Cost per performance in AI seems to be falling more than ~200x annualized (and higher at higher levels of benchmark performance.) And exceeding 500x annualized cost declines moving along the efficient frontier of the performance curve. Probably more relevant practically for users, at $1 per task whereas today you might only have 50/50 odds that the agent succeeds, by the end of year, you should have an 89% chance of success and by a year from now a 97% chance (at least on deepSWE 1.1 type tasks.) The moving pareto frontier makes it difficult to cleanly report a performance cost improvement due to the shape of the cost performance curve. At the highest asymptote of performance you go from literally not being able to achieve such a low error rate *at any cost* to being able to get that performance for 10s of $s per task (an infinite cost decline). We forecast that top-end rate separately. In the belly of the curve you need only cross a performance cost threshold, but as the curve steepens you can also less expensively buy more performance (though not 100% clear that the curve really is steepening that much.) We also model that performance curve, though given uncertainty effectively zero out continued improvement for the conservative case. And at lower levels of benchmark performance you get a cleaner understanding of the underlying cost-per-performance improvement though at thresholds that are less interesting practically. If anything I suspect that people are wildly under indexing on the rate of improvement of these models. Even last month's AI spend will fall to a fraction of a fraction within the year (if we weren't going to so predictably deploy the new capabilities against new workloads, tasks and challenges.) Spending a lot of time trying to painstakingly carve out specific current workflows into more efficient models, at this point in the improvement curve, seems, if anything, a fool's errand designed to line the pockets of a consultant. Let the tokens flow.

Thoughts on the "ChatGPT + Codex = ChatGPT Work" app Some background, I've been using our internal Slack bot for the past couple of weeks and I feel convinced that Claude Cowork and ChatGPT Work are the wrong type of product. 1. People will naturally use an agent where they already do work (Slack or Teams), as opposed to building a habit around a new app. 2. The agent is only as good as the tools and data it has access to, and there's quite a bit of friction in setting that up. It's way better for somebody on your team to do that once, as opposed to, every single person having to do that for their desktop agent. 3. The desktop agent has a clear advantage though with computer use. And while that is powerful, most knowledge work is now done in the cloud. So oddly the desktop agent is better for people that mostly do their work on their local machine, ie, developers. On that note, merging Codex and ChatGPT to create ChatGPT Work feels like the wrong move on multiple counts. It dilutes two previously focused products (ChatGPT and Codex) to create one convoluted app, that I don't think is the right form factor to begin with.