journalfront

The Creator of ChatGPT's Voice Aims to Build Tech from "Her" Without Dystopia

2024-12-09

Alexis Conneau has long been fascinated by the movie "Her" and has been dedicated to turning its fictional voice technology into reality. His Twitter banner features a picture of Joaquin Phoenix's character from the movie. With ChatGPT's Advanced Voice Mode, a project he started at OpenAI after similar work at Meta, he achieved a significant milestone. Now, he has launched WaveForms AI, a new audio LLM company aiming to release AI audio products in 2025 that compete with those from OpenAI and Google. The startup raised $40 million in seed funding led by Andreessen Horowitz.

How "Her" Influenced Conneau and His New Venture

In an interview with TechCrunch, Conneau spent a good deal of time discussing how to avoid the dystopia depicted in "Her". The movie was a science fiction tale about a world where people form intimate relationships with AI systems instead of humans. Conneau believes they want to bring the existing and future technology for good and do the opposite of what the company in the movie does. Building the tech without the dystopia seems contradictory, but he is determined to do it and is convinced his new AI startup will help people "feel the AGI" with their ears.WaveForms AI is an audio LLM company training its own foundation models. It aims to supply "emotionally intelligent" AI that facilitates various interactions, such as talking to your car or computer. Conneau is wary of the AI companionship space but wants the company to be more "horizontal" and have a broader impact. He believes talking to generative AI will become a more common way to interact with technology in the future.

The Technology Behind WaveForms AI

WaveForms AI's Advanced Voice Mode is a significant innovation. It doesn't just translate voice into text and back like the old voice mode. Instead, it breaks down the audio of your voice into tokens and runs those tokens directly through an audio-specific transformer model, enabling low latency. This is a more advanced approach that sets it apart from ChatGPT's regular voice mode.When it comes to AI audio models, there is a claim that they can "understand emotions". Similar to text-based LLMs, audio LLMs recognize patterns in audio clips labeled as "sad" or "excited" by humans. They respond with emotional intonations based on these recognized patterns.

Making AI More Personable

Conneau believes that generative AI doesn't need to get significantly smarter than GPT-4o to create better products. WaveForms is focusing on making AI better to talk to rather than improving the underlying intelligence. They aim to develop smaller foundational models that are less expensive and faster to run, given the evidence that the old AI scaling laws are slowing down.Conneau's former co-worker at OpenAI, Ilya Sutskever, often talked about trying to "feel the AGI". Conneau is convinced that achieving AGI will be more of a feeling and that audio LLMs will be the key. He believes you will be able to feel the AGI more when you can talk to it and hear it.

Responsibility in AI Development

As startups make AI better to talk to, they have a responsibility to ensure people don't get addicted. Andreessen Horowitz's Martin Casado believes it may not be a bad thing if people talk to AI more often. He compares it to talking to a random person on the internet or a video game. However, from a societal standpoint, there is a concern about developing a loving relationship with AI, as depicted in "Her". WaveForms now has to navigate this fine line.

It seems that users within the EU and UK will be denied access to OpenAI's highly anticipated Sora video generation model upon its initial launch. This significant development has sparked curiosity and concern among tech enthusiasts and industry observers alike. A recently published help page on OpenAI's website clearly shows that the supported regions for Sora "on web and mobile" do not include any EU countries. On this page, OpenAI has explicitly stated that attempting to access Sora outside the listed territories may lead to the suspension or banning of accounts.

Unveiling the Exclusion: OpenAI's Sora and the EU/UK Dilemma

Section 1: The Initial Exclusion

1: The news of OpenAI's Sora being unavailable in the EU and UK at launch has sent ripples through the tech community. This exclusion raises questions about OpenAI's global strategy and its commitment to providing equal access to its innovative technologies. It is a notable setback for users in these regions who were eagerly anticipating the arrival of Sora. 2: The newly published help page serves as a clear indication of OpenAI's decision. By omitting all EU countries from the list of supported regions, the company has effectively shut the door on potential users in these areas. This move has drawn attention to the challenges and complexities of operating in a global marketplace with diverse regulatory environments.

Section 2: Past Precedents

1: This is not the first time OpenAI has faced such a situation with an initial product launch. Earlier this summer, when the company began rolling out its Advanced Voice Mode, a human-like conversational feature for ChatGPT, EU users were left out of the early waves. This pattern suggests that there may be underlying issues or considerations that lead to the exclusion of EU countries in certain product launches. 2: In a statement provided to TechRadar this fall, OpenAI attributed the Advanced Voice Mode delay to the "additional external reviews" required by some territories. The company emphasized that this is a common practice to ensure that their features align with local requirements. However, the prolonged waiting period for EU users has raised concerns about the efficiency and fairness of these review processes.

Section 3: Impact on Other Tech Companies

1: Other tech companies working on AI models and products, such as Meta and Microsoft, have also encountered similar challenges in the EU due to the bloc's complex data privacy regulations. These regulations have forced these companies to push back their AI product releases, highlighting the significant impact that regulatory frameworks can have on the development and deployment of AI technologies. 2: Meta, in particular, has been vocal about the onerous compliance requirements it faces. Earlier this year, Meta endorsed an open letter calling for a "modern interpretation" of European privacy laws that does not "reject [AI] progress." This indicates the ongoing struggle that tech companies face in balancing innovation and compliance in the EU.

Artificial intelligence is making waves in the field of breast-screening, with one U.S. program showing remarkable results. Women who chose AI-enhanced scans saw a 21% higher likelihood of cancer detection. DeepHealth, a radiology firm, presented these findings at the RSNA annual meeting. They analyzed 747,604 women's mammography screenings over 12 months. The overall cancer detection rate was 43% higher for those in the program. Some women also used FDA-compliant AI software at their own expense to assist radiologists in spotting anomalies.

Unraveling the Impact of AI in Breast-Screening

The Role of Selection Bias

The findings suggest that 22% of the cancer detection increase might be due to higher-risk patients enrolling first, indicating potential selection bias. This highlights the need for careful consideration when evaluating the benefits of AI in breast-screening. It shows that simply looking at the overall numbers may not tell the whole story.

For example, if only those with a higher risk of cancer are more likely to choose the AI-enhanced scans, it could skew the results. This aspect needs to be taken into account when assessing the true value of AI in mammography.

The Contribution of AI in Driving Recall Rates

The remaining 21% of the cancer detection increase was attributed to the use of AI in driving recall rates for additional imaging. This demonstrates the significant role that AI can play in identifying potential issues that might otherwise be overlooked.

By providing a "second set of eyes," AI helps radiologists focus on areas that require further examination. This not only leads to more accurate cancer detections but also potentially saves lives by catching cancers at an earlier stage.

The Need for Randomized Controlled Trials

Despite the promising initial results, the research is far from conclusive. The researchers are now working on randomized controlled trials to better quantify the benefits of AI in this context.

This is a crucial step as it will provide more reliable data and help determine the true impact of AI on breast-screening. Without such trials, it will be difficult to make definitive claims about the effectiveness of AI in this field.