I see you as a founder regardless @huw. It's a weird word but it's well accepted in the social enterprise works at least that people who come in 2-3 years down the line and creatively grow an org are founders. I suppose it's just semantics though.

I've been mulling over this quote from Naomi Klein over the last couple of days. I think its a strong summary of one of the best ethical arguments against the top AI labs. 

My argument against this might be that the actual purpose of commercial application is to improve human wellbeing and prosperity overall, not to eliminate jobs. Jobs may or may not be eliminated, but either option could be fine if the prosperity is shared (at least somewhat) throughout humanity. 

Then there are orgs like Mechanize, which are explicitly trying to eliminate jobs...

Besides that on the "theft" of creativity front, I think this is broadly true but I'm not sure what can be done at this point. To generalise (even with coding) AI feeds of the best that humanity has to offer then produces worse-than-the-best output much faster, at a fraction of the cost. Without the best of human IP, AI wouldn't be very good. Newer models may be starting to be better than the best humans in niche areas, but this isn't the norm.

I talk a lot about how AI helps us provide healthcare to some of the poorest people, but I still don't have the greatest response to these kind of criticisms from many of my friends. I wonder how others respond to people when they bring arguments like this?
 

Love the curating good posts from years ago approach. This one was great, keep it up.

Want to add that the writer Abi has been great and responded really well to feedback.

This is a super important cause, but I think these numbers are hugely overblown.

That 125,000 to 250,000  deaths following blindness an old figure from the 90s, deaths from vitamin A deficiency have hugely dropped since then. 

I think also from 2010 to 2015 golden rice was also a little lower yielding which contributed to the lack of uptake along with the GMO vitriol? So uptake was never going to be overwhelming until well after 2015 I don't think bans nonwithstanding

GBD estimated around 17,000 deaths from VAD in 2021
https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2025.1689903/full

Its a nice idea for a counter, but might be like 3-10x off or something? Have messaged the author directly.
 

Quick question - how bent towards AI safety work is the book? I'm keen to recommend it but I'm not interested in recommending or buying it for a general audience if it pushes AI safety harder than other cause areas (even if not much more strongly). Mainly because this would just turn many people I might recommend it to off. Like most people, they just would never be interested in working in AI safety.

From the blurb this looked positive "The overall structure of the book remains the same – it’s still a cause-agnostic introduction to what makes for a satisfying job, with classic EA advice on how to have an impact, how to build skills, and how to make a career plan. AI is discussed where it’s relevant to these questions (see the table of contents in the FAQ below)."

But then this made me think again

The book covers the possibility of short AI timelines and what they mean for your career, but we don’t think they’re certain, and the book is about careers first, rather than short AI timelines (80k has a bunch of separate material on career advice for mitigating AI risk).

And I wonder what the conclusion of this chapter "What’s the World’s Most Pressing Problem?" is.

I'm definitely not saying its bad if it has a clear AI bent, but was wondering what people's thoughts on this is in reference to my target audience of mid 20s to early 30s people who I know won't be interested in working on AI risk.

One thing I really like about founding to give programs, is that we have very clear metrics for success. We'll know over 2-5 years if these programs were successful or not. 

I haven't seen the success metrics, but I'm guessing It probably only takes 1 or 2 successful companies to make the whole thing worth it and I'm hoping the cohorts can do even better than this!

Thanks for this interesting look. I strongly agree with this sentiment, and probably havent' emphasised it enough myself in this AI discourse.

  "Writing is thinking. I predict we will think less clearly when we stop writing ourselves and start outsourcing it to our ever-willing country of editors."