🗣️ AI voice cloning

PLUS: China vs USA - who will win the AI race?

New here? Welcome to NAB, and make sure to subscribe to stay informed on the latest developments in AI!

Happy Tuesday, fellow humans 👋

OpenAI released a new voice cloning tool they are too scared to release to the public, bots might be better at debating than humans, and China is trying its hardest to create an OpenAI competitor…

Let’s dive into it!

Have anything you’d like to share with over 40k AI enthusiasts?

🧵 In today's edition:

  • 🗣️ AI voice cloning

  • 🧠 The power of persuasion

  • 📈 China vs USA - who will win the AI race?

  • 🤑 AI Fundraising News

  • 🗞️ AI Quick-Bytes

  • 🐦 Tweet Post of the Day

Check out the full demo here!

Over the past year, deepfakes have become a growing concern, often causing confusion between reality and fiction.

In a sea of voice-generation tools, OpenAI's new Voice Engine tool has taken this to a whole new level, generating incredibly realistic voice clones with just 15 seconds of audio.

But as tempting as it sounds to talk like Morgan Freeman, OpenAI is hitting the pause button on its worldwide release, deeming the tool too risky amidst this year's pivotal global elections.

Here's why this is a wise decision:

  • The risk of spreading misinformation is high, prompting OpenAI to adopt a "cautious and informed" approach.

  • Plus, the tool is currently only available to select partners who have obtained explicit consent from the original speakers.

However, the demonstrated applications are intriguing, ranging from educational voiceovers to restored voices lost to illness.

While OpenAI advocates for phasing out voice authentication for security, competitors like ElevenLabs already offer public voice cloning using minutes of audio. However, their "no-go voices" filter aims to prevent the imitation of active political candidates.

So, how can we prevent this technology from becoming uncontrollable? It might be vital to discuss ethics before synthetic voices become the next deepfake nightmare.

Read more: The Guardian

An interesting study revealed that language models can now sway opponents in debates more than humans, even when individuals know they're engaging with an AI.

Process: Researchers matched 820 participants against both human and AI-powered opponents, such as GPT-4, in a series of animated debates on contentious topics.

The result? Remarkably, participants were 81.7% more likely to be persuaded by the AI's arguments when it had access to their personal details.

So, what gives these chatbots such formidable powers of persuasion?

  • Their extensive knowledge banks allow them to reference obscure facts that typical humans can't access.

  • The AI also remained refreshingly impartial, free from any hidden agenda.

However, some experts are skeptical whether these controlled experiments genuinely reflect the complexity of actual human discourse. After all, being coerced to argue for a stance you don't support is quite different from a sincere conflict of opposing viewpoints.

Regardless, the implications are immense. As language AIs become more sophisticated, they may acquire startling abilities to alter our perspectives through pure rhetorical powers.

I may need to re-read Influence at some point…

Read more: New Scientist

While pioneers like OpenAI, Google, and Microsoft are leading the AI race with advanced Language Models (LLMs) and substantial funding, China is working hard to catch up.

How? By utilizing open-source LLMs like Meta's Llama 1, Chinese giants such as Baidu, Alibaba, and Tencent are quickly creating their own ChatGPT competitors.

How’s the progress?

  • It’s looking like China's leading contenders might still be 1-2 years behind OpenAI's latest advancements.

  • However, they are catching up quickly, propelled by a large talent pool that surpasses the U.S. in producing top-tier AI graduates.

Plus, the AI tech cold war has complicated China's plans:

  • U.S. trade restrictions limit Beijing's access to vital AI chips produced by companies like Nvidia.

  • Nevertheless, Chinese frontrunners like Huawei are intensifying their efforts to produce homegrown semiconductors, aiming to reduce dependence on American silicon.

All of this is to say: this race is FAR from over 😤

Read more: CNBC

🤑 AI Fundraising News

  • Brandtech Group raises $115M in funding to develop a GenAI-based marketing agency.

  • Cyvl.ai raises $6M in funding to utilize vehicle-mounted 3D mapping sensors and AI algorithms to enhance how engineers inspect and manage physical infrastructure to improve efficiency, reduce costs, and improve safety.

🗞️ AI Quick-Bytes

What else is going on?

🐦 Tweet Post of the Day

Why is this so good tho 😂 

What did you think of today's newsletter?

Login or Subscribe to participate in polls.

And that does it for today's issue.

As always, thanks for reading. Have a great day, and see you next time! ✌️

— Haroon: (definitely) Not A Bot and @haroonchoudery on Twitter

P.S.: Check us out on our socials - snacks on us if you also hit follow. :)

Join the conversation

or to participate.