Hi there,

Welcome to Impact Statement, a newsletter delivered once per week, that takes two-minutes to read, with three useful sections — Thoughts, Resources, and Grantable News.

Thoughts 🧠

Same amount of good judgement as before

It occurred to me recently that the amount of good judgement it takes to write a great grant proposal is exactly the same as before generative AI changed the world.

While a ten-hour grant proposal might only take you one hour to draft using AI assistance, you still need to use just as much good sense as before (maybe more?) to guide the project.

AI misinformation is increasingly harmful

In the aftermath of Hurricane Helene, we've seen inflammatory AI generated imagery spreading online, leading to problems for people working to solve urgent problems on the ground.

There are efforts underway to mandate things like watermarks and disclaimers that acknowledge the use of AI to protect consumers (see below), but it's also the case that information literacy is going to become an increasingly important skill.

AI's improved reasoning abilities

Lastly, I wanted to share that AI models have taken some large steps forward recently, particularly in their abilities to reason.

With the release of OpenAI's GPT-o1 model, we are beginning to see the fruits of efforts to solve for this vulnerability. When given International Math Olympiad questions as a test, GPT-4o, the standard ChatGPT model scored 14% correct, while GPT-o1 achieved 84% correct responses.

"OpenAI o1 works by using chain-of-thought logic; it talks itself through a series of reasoning steps as it attempts to solve a problem, correcting itself as it goes." Nature, article link

For many who may have witnessed AI stumbling over simple reasoning problems and rolled your eyes, this is a reminder that the field of AI is extremely dynamic and improving rapidly.

Resources 📝

Here's what early AI legislation is shaping up to look like:

California bill SB 942

"Existing law requires the Secretary of Government Operations to develop a coordinated plan to, among other things, investigate the feasibility of, and obstacles to, developing standards and technologies for state departments to determine digital content provenance. For the purpose of informing that coordinated plan, existing law requires the secretary to evaluate, among other things, the impact of the proliferation of deepfakes, defined to mean audio or visual content that has been generated or manipulated by artificial intelligence that would falsely appear to be authentic or truthful and that features depictions of people appearing to say or do things they did not say or do without their consent, on state government, California-based businesses, and residents of the state."

Want to see how good AI-generated imagery has become?

Reddit thread

Read more about how GPT-o1 exceeds PhD-level performance

‘In awe’: scientists impressed by latest ChatGPT model o1, Nature

"Strikingly, o1 has become the first large language model to beat PhD-level scholars on the hardest series of questions — the ‘diamond’ set — in a test called the Graduate-Level Google-Proof Q&A Benchmark (GPQA). OpenAI says that its scholars scored just under 70% on GPQA Diamond, and o1 scored 78% overall, with a particularly high score of 93% in physics."

Grantable news 📰

Check out Part 2 with Holly Rustick!

Our first Grantable Partner 🤝 People and products we believe in sharing with our community.

Don't miss Part 2 of our interview with Holly Rustick, founder of Grant Writing And Funding and a best-selling author, to learn more about Holly’s upcoming Freelance Grant Writer Academy — a 12-month group coaching program where aspiring and seasoned grant writers replace their full-time income with writing grants from home.

Want to learn more?

Tuesday, Oct 15th at 4pm - 5:30 pm EST

Meet Holly and learn about how her students have broken free of professional constraints to create flexible and financially fulfilling careers as freelancers.

Thanks for reading this edition of Impact Statement and stay tuned for more.