🍕 There’s no way out

Hey there,

I’m back from holiday and immediately had 3 crazy-packed days! In the past 3 days, I spoke at a conference in Slovakia, one in Italy for the Ministry of Culture, and one at the AI Innovation House in Denmark. There’s so much energy, what a time to be part of this revolution ❤️

Now let’s get to this week’s news:

• There’s no way out

• Pizza Bytes 🍕: Welcome DALL·E 3, Google Bard update, EU supercomputers to AI startups

• Tiramisù 😋: OpenAI images choice for DALL·E 3 presentation is amazing

Ok let’s get started 🕺

There’s no way out.

There are still a lot of people who wonder whether AI is really going to completely change our lives or not.

Let me talk about a recent study done by Harvard with Boston Consulting Group (BCG). The study is called "Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of AI on Knowledge Worker Productivity and Quality”, I recommend you read the whole thing here, because there are so many interesting things there but today I’ll focus on a single aspect: productivity.

The experiment involved 758 consultants who had to complete tasks they normally do in their wildly high-paid jobs, and were randomly assigned to one of three conditions: no AI access, GPT-4 AI access, or GPT-4 AI access with a prompt engineering overview.

I’ll cut to the chase and quote two facts:

• “Consultants using AI were significantly more productive (they completed 12.2% more tasks on average, and completed tasks 25.1% more quickly), and produced significantly higher quality results (more than 40% higher quality compared to a control group).”

• Those below the average performance threshold increasing by 43% and those above increasing by 17% compared to their own scores.

These are not random numbers from a silly test writing silly copy for marketing ads. These are complex tasks that VERY high-paid people at a VERY selected firm do. A senior-level executive at the company commented on these tasks being “very much in line with part of the daily activities”. In a capitalistic society, who would say “no” to a 40% quality gain with a 25% cut in time?

No one who is mentally OK.

This is just another data point that goes in the same direction: there is NO doubt that AI will force us to completely rethink the way we work, and potentially the whole fabric of our society (what happens when anyone has access to the same skills of the top world’s professionals for the price of a pizza per month?).

So what do we need to do?

I have one advice, and two reasons for it. My advice is GET INVOLVED, don’t watch this massive change from the sidelines. I have two reasons for it:

1. It’s a massive career change. The people who master AI today will be the leaders of tomorrow.

2. We need to make sure that we get this right, that everyone is represented and no one is left behind. And we need more people to study, understand, and use AI. Join us!

Pizza Bytes 🍕

• OpenAI has introduced DALL·E 3, an improved version of its image-generating AI system. DALL·E 3 can create more accurate and detailed images based on specific prompts, and it is designed to decline requests for violent or hateful content. Also: it’s integrated into ChatGPT!

• Google has launched an update to their conversational AI system, Bard, which integrates with Google apps and services for more helpful responses. The update also includes features like double-checking answers using Google search and collaborating on documents and conversations.

• OpenAI has published its Risk Reduction Strategy (RSP) which outlines the framework for addressing catastrophic risks associated with AI models. The RSP includes an AI Safety Levels (ASL) system that defines safety measures and standards to mitigate risks while incentivizing safety progress.

• Director Gareth Edwards experimented with AI-generated music for the soundtrack of his movie but ultimately chose renowned composer Hans Zimmer. The AI-generated track was deemed only a 7 out of 10 “but the reason you go into Hans Zimmer is for 10 out of 10,” says Gareth Edwards emphasizing the importance of human creativity and taste in art.

• Anthropic has partnered with Boston Consulting Group to bring its AI assistant, Claude, to more enterprises. BCG will use Claude to synthesize research, analyze data, and provide insights to clients, aiming to set a new standard for responsible enterprise AI.

• Researchers at DeepMind have developed AlphaMissense, a powerful tool that predicts the harmfulness of certain genetic variants in the human genome. By using a combination of structural analysis and evolutionary information, this model achieves impressive accuracy.

• The European Union plans to expand access to its high-performance computing supercomputers for AI startups to train their models. However, startups will need to comply with the EU's AI governance program, which includes regulations outlined in the AI Act.

• The UK government is launching a pilot scheme that will allow businesses to demonstrate their compliance with regulatory standards for AI and digital innovations. The scheme aims to provide tailored support and streamline the process for businesses to bring their innovations to market quickly and responsibly.

• A group of artists using generative AI tools have penned an open letter advocating for a more inclusive and informed discussion around AI in the arts. They argue that generative AI tools empower artists, allow for new transformative works, and urge lawmakers to consider their perspective in shaping the development and regulation of AI.

• Microsoft's GitHub repository had a misconfigured link that granted unauthorized access to private data, including personal computer backups of Microsoft employees. The account contained sensitive information such as passwords, secret keys, and internal messages.

Tiramisù 😋

The sweet part of the newsletter: fun stuff from the crazy tech world.

• Let’s stop for a moment to appreciate the images chosen by OpenAI to present DALL-E 3

You reached the end of this edition 😢

I’ll talk to you next week.

Ciao 👋