OpenAI is reportedly poaching Google's top AI researchers with $10m/year packages.

The recruiting fight between OpenAI and Google is growing fiercer.

As OpenAI proceeds with an employee share sale that would nearly triple the startup’s valuation to more than $80 billion, its recruiters are courting top artificial intelligence employees at Google with millions of dollars and a message: Join us now to lock in a stock package at the current valuation of $27 billion and benefit from the impending increase.

As part of their pitch, OpenAI recruiters have claimed researchers would have greater access to computing resources, including the specialized chips staff rely on to run experiments and develop new techniques for AI models, according to people with knowledge of the claims.

Financial Times: Study Shows that ChatGPT boosts low-performing employees the most

This Financial Times article discusses the impact of generative artificial intelligence, like OpenAI's ChatGPT and GPT-4, on white-collar jobs. Key points include:

1. Risk to High-Wage Jobs: OpenAI estimates that higher-paying jobs are more at risk from AI advancements. Those earning six-figure salaries are three times more likely to be affected than those earning $30,000.

2. Immediate Impact on Freelancers: A U.S. study found that after the launch of ChatGPT, freelancers in fields like copywriting and graphic design experienced a decrease in jobs and earnings. This suggests that AI is not only taking over some tasks but also devaluing the remaining work.

3. Varied Impact Across Skill Levels: The most skilled freelancers were not shielded from the negative impact. In fact, they often faced worse outcomes.

4. Productivity and Quality Increases with AI: A Harvard Business School study involving Boston Consulting Group employees showed that those using GPT-4 were more productive and produced higher quality work than those without AI assistance. The largest gains were observed among less skilled workers.

5. AI's Limitations and Human Collaboration: GPT-4 struggled with nuanced tasks requiring careful analysis of qualitative materials. However, employees who effectively integrated AI into their workflow ("cyborgs") or focused on tasks suited to their expertise while using AI for others ("centaurs") performed better.

6. Implications and Recommendations: The article suggests that regulation is crucial, particularly in unregulated markets like online freelancing. Roles involving multiple facets are less at risk of automation. Finally, maximizing the benefits of AI tools requires treating them as extensions of ourselves, actively engaging with and scrutinizing their outputs.

MIT Found the same thing earlier this year regarding call center agents:

We've been flying blind until now

Cool demo of OpenInterpreter writing the code for a UI just by looking at it using the new ChatGPT Vision API

A new version of GPT-4 Turbo is now live in ChatGPT

Other stuff

• Exceptionally good overview of adversarial attacks on LLMs.

• Introducing, Motion Brush by Runway

• AI could cause ‘catastrophic’ financial crisis, says Yuval Noah Harari

• Tutorial: Text-to-speech on Linux/Mac using Bash made by ChatGPT 4

• How to Stop Feeding AWS’s AI With Your Data

Prompt of the Day 🍭: The AIDA Formula (Attention, Interest, Desire, Action)

Click on the image to load the full prompt in ChatGPT (requires Superpower ChatGPT)

Hope you enjoyed today's newsletter

⚡️ Join over 200,000 people using the Superpower ChatGPT extension on Chrome and Firefox.