OpenAI Used Workers on $2 Per Hour to Make ChatGPT Less Toxic 

OpenAI used outsourced Kenyan laborers earning less than $2 per hour, a TIME investigation has found.

How OpenAI did it:

They scraped content about topics they didn't want ChatGPT talking about (e.g., murder, self-harm, sexual abuse, incest)

They sent it to Sama AI along with a $12/hr contract to get that content labeled

Sama AI paid Kenyan workers $2/hr to read and label 150-250 passages each nine-hour shift. 

This Investigation gets crazy quick and super long so;

Read the full story 

Read longer summary 

An AI just passed an Economics university exam 

“Claude” — a similar AI system designed by SBF-financed Anthropic — has now passed a law and economics exam. Alex Tabarrok, an economics professor at George Mason University, writes on the Marginal Revolution blog he runs with Tyler Cowen that Claude earned a “marginal pass” on a recent blind-graded test. 

Alex also commended Claude for being able to give some answers better than Humans, especially when asked how to improve the law and economics of intellectual property.

Read More 

5 highlights from Sam Altman's interview 

Note: We are just summarizing what he said with a bit of humor not word for word

On the release of GPT-4: 

Hold your horses' folks, GPT-4 is coming, but we want to make sure we do it safely and responsibly. Don't get too excited, we're gonna take our time before releasing it. But trust us, eventually, you'll be glad we did.

On the expectations for GPT-4: 

Looks like people are setting themselves up for disappointment. The hype is real but AGI is not. Sorry to say, but we'll probably disappoint some folks. To be honest we all have extremely granular expectations of GPT 4 maybe we should all just take a chill pill

On the unexpected progress of AI: 

Everyone thought machines would first take over physical labor, then less demanding jobs, and lastly creative jobs. But looks like it's going in the opposite direction, with machines tackling more cognitive tasks, and creativity being the last to go or maybe never. Who knows, maybe there's something special about human creativity that machines can't replicate.

On ChatGPT being integrated with Microsoft Office: 

You are a very experienced and professional reporter. You know I can't comment on that. I know you know I can't comment on that. You know I know you know you can't comment on that. In the spirit of the shortness of life and our precious time here, why do you ask?This is sams actual reply and I'm dying of laughter

On Google building an AI: 

I haven't seen google's work, but I assume they're good. But don't count out Google just yet. People always think technology will be the end of big companies, but they tend to forget that these companies can adapt. There might be a shift in search, but it's not going to happen overnight. Google will probably still be the go-to search engine for a while. And whatever this "code red" thing is, it's probably not going to change that much.

Read More

The NHS using AI to help Patients

An AI chatbot that can predict mental health disorders with 93% accuracy has become the first in the world to achieve Class IIa UKCA medical device certification. This certification confirms that the tool is clinically effective, safe, and can be safely incorporated into the psychological therapy pathway to support patient self-referral. This will help overstretched clinicians and improve patient outcomes. The tool can classify the 8 common mental health disorders treated by NHS Talking Therapies and has been used by 130,000 NHS patients so far.

Read More