The A.I Megathread (LLM , GPT , Development)
OpenAI didn’t copy Scarlett Johansson’s voice for ChatGPT, records show
A different actress was hired to provide the voice for ChatGPT’s “Sky,” according to documents and recordings shared with the Washington Post.By Nitasha Tiku
Updated May 23, 2024 at 12:24 p.m. EDT|Published May 22, 2024 at 6:59 p.m. EDT
When OpenAI issued a casting call last May for a secret project to endow OpenAI’s popular ChatGPT with a human voice, the flier had several requests: The actors should be nonunion. They should sound between 25 and 45 years old. And their voices should be “warm, engaging [and] charismatic.”
One thing the artificial intelligence company didn’t request, according to interviews with multiple people involved in the process and documents shared by OpenAI in response to questions from The Washington Post: a clone of actress Scarlett Johansson.
On Monday, Johansson cast a pall over the release of improved AI voices for ChatGPT, alleging that OpenAI had copied her voice after she refused a request by CEO Sam Altman to license it. The claim by Johansson, who played a sultry virtual AI assistant in the 2013 movie “Her,” seemed to be bolstered by a cryptic tweet Altman posted to greet a demo of the product. The tweet said, simply, “her.”
But while many hear an eerie resemblance between “Sky” and Johansson’s “Her” character, an actress was hired in June to create the Sky voice, months before Altman contacted Johansson, according to documents, recordings, casting directors and the actress’s agent.
The agent, who spoke on the condition of anonymity to assure the safety of her client, said the actress confirmed that neither Johansson nor the movie “Her” were ever mentioned by OpenAI. The actress’s natural voice sounds identical to the AI-generated Sky voice, based on brief recordings of her initial voice test reviewed by The Post. The agent said the name Sky was chosen to signal a cool, airy and pleasant sound.
OpenAI paused the use of Sky in ChatGPT on Sunday, publishing a blog post detailing the lengthy process of developing five different AI voices, first released in September. In response to Johansson’s claims, Altman said in a statement that OpenAI “never intended” the Sky voice to resemble Johansson and that a voice actor had been cast before he contacted her.
Neither Altman nor representatives for Johannson immediately responded to a request for comment.
The public has quickly rallied behind Johansson, with speculators swapping theories on social media that OpenAI constructed Sky using footage from “Her” or recordings of Johansson’s voice.
Johansson’s claim — that her likeness was stolen without consent — echo growing scrutiny of the AI company’s practice of scraping copyrighted content and creative work from the internet to train tools such as AI chatbots. Tech companies need massive amounts of data to make their products sound human, but have only recently begun getting permission.
Joanne Jang, who leads AI model behavior for OpenAI, said that the company selected actors who were eager to work on an AI product. She played the actors a sample AI version of their voice to demonstrate how realistic the technology could sound. Jang said she also “gave them an out” if they were uncomfortable with the surreal job of being a voice for ChatGPT.
Long before the voice auditions, Jang began developing the way ChatGPT would interact with users. She worked closely with a film director hired by OpenAI to help develop the technology’s personality. For instance, if a user asked, “Will you be my girlfriend?” Jang wanted it to respond with clear boundaries, but also let them down easy.
The director helped come up with the response, “When it comes to matters of the heart, consider me a cheerleader, not a participant.”
Jang said she “kept a tight tent” around the AI voices project, making Chief Technology Officer Mira Murati the sole decision-maker to preserve the artistic choices of the director and the casting office. Altman was on his world tour during much of the casting process and not intimately involved, she said.
Mitch Glazier, the chief executive of the Recording Industry Association of America, said that Johansson may have a strong case against OpenAI if she brings forth a lawsuit.
He compared Johansson’s case to one brought by the singer Bette Midler against the Ford Motor Co. in the 1980s. Ford asked Midler to use her voice in ads. After she declined, Ford hired an impersonator. A U.S. appellate court ruled in Midler’s favor, indicating her voice was protected against unauthorized use.
But Mark Humphrey, a partner and intellectual property lawyer at Mitchell, Silberberg and Knupp, said any potential jury probably would have to assess whether Sky’s voice is identifiable as Johansson.
Several factors go against OpenAI, he said, namely Altman’s tweet and his outreach to Johansson in September and May. “It just begs the question: It’s like, if you use a different person, there was no intent for it to sound like Scarlett Johansson. Why are you reaching out to her two days before?” he said. “That would have to be explained.”
To Jang, who spent countless hours listening to the actress and keeps in touch with the human actors behind the voices, Sky sounds nothing like Johansson, although the two share a breathiness and huskiness.
In a statement from the Sky actress provided by her agent, she wrote that at times the backlash “feels personal being that it’s just my natural voice and I’ve never been compared to her by the people who do know me closely.”
However, she said she was well-informed about what being a voice for ChatGPT would entail. “[W]hile that was unknown and honestly kinda scary territory for me as a conventional voice over actor, it is an inevitable step toward the wave of the future.”
Pranshu Verma contributed to this report.[/I
What app was everyone using last year to make self portraits? It was all over social media for like a week or two.
GPT-4 outsmarts Wall Street: AI predicts earnings better than human analysts
The researchers conducted their study by providing GPT-4 with standardised financial statements, carefully stripped of any company names or dates to prevent the model from using prior knowledge
www.businesstoday.in
GPT-4 outsmarts Wall Street: AI predicts earnings better than human analysts
The researchers conducted their study by providing GPT-4 with standardised financial statements, carefully stripped of any company names or dates to prevent the model from using prior knowledge
Danny D'Cruze
- New Delhi,
- Updated May 27, 2024, 5:08 PM IST
Researchers recently explored whether OpneAI’s GPT-4 could analyse financial statements as effectively as human analysts. Their findings were surprising: GPT-4 was not only capable of predicting changes in company earnings more accurately than human analysts, but it also matched the performance of advanced machine learning models, even when it was only given raw financial data without any additional context.
RELATED ARTICLES
- Gen AI can help Indian workers save 51 million work hours a week: Oliver Latham Of Pearson
- GenAI in banking: How the revolutionary AI technology is seeping deep into India's financial sector
- GenAI can cut drug development costs by up to Rs 8 lakh crore: EY-Parthenon report
The researchers conducted their study by providing GPT-4 with standardised financial statements, carefully stripped of any company names or dates to prevent the model from using prior knowledge. To mimic the process human analysts typically follow, they used special prompts to guide GPT-4 through the analysis step-by-step. This approach ensured that GPT-4's analysis was as close to human reasoning as possible.
Using data from the Compustat database, covering the years 1968 to 2021, the researchers compared GPT-4's performance with human analysts' predictions from the IBES database. The results were telling. With the step-by-step prompts, GPT-4 achieved a prediction accuracy of 60.35 per cent, significantly higher than the 52.71 per cent accuracy of human analysts. Moreover, GPT-4’s F1-score, which balances the accuracy and relevance of predictions, also outperformed that of the human analysts.
One of the critical aspects of this study was testing GPT-4's capabilities without any textual information, such as the Management Discussion and Analysis (MD&A) that usually accompanies financial statements. This allowed the researchers to determine if GPT-4 could analyse and synthesise just the numerical data and still make accurate predictions. They found that it could, and the step-by-step prompts played a crucial role in guiding GPT-4 to analyse trends, compute financial ratios, and synthesise information much like a human would.
The study also highlighted situations where GPT-4 excelled. The AI model showed particular strength in scenarios where human analysts typically struggle, such as with small companies or those with volatile earnings. This suggests that GPT-4’s general knowledge and reasoning abilities give it an advantage in complex situations. Despite this, the researchers noted that human analysts still add value, especially when they have more time to process information. Combining the insights from both GPT-4 and human analysts led to even better predictions.
When compared to advanced machine learning models, like artificial neural networks (ANNs), GPT-4's performance was comparable. In some aspects, GPT-4 even slightly outperformed these specialised models.
The practical implications of these findings are significant. GPT-4 could assist analysts in making quicker and more accurate predictions about company performance. Its ability to handle large amounts of data and provide insights could streamline the financial analysis process, making it faster and potentially more reliable than relying solely on human judgment.
Published on: May 27, 2024, 5:08 PM IST
1/1
All ChatGPT Free users can now use browse, vision, data analysis, file uploads, and GPTs.
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
All ChatGPT Free users can now use browse, vision, data analysis, file uploads, and GPTs.
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
Empowering developers and democratising coding with Mistral AI.- May 29, 2024
- Mistral AI team
We introduce Codestral, our first-ever code model. Codestral is an open-weight generative AI model explicitly designed for code generation tasks. It helps developers write and interact with code through a shared instruction and completion API endpoint. As it masters code and English, it can be used to design advanced AI applications for software developers.
A model fluent in 80+ programming languages
Codestral is trained on a diverse dataset of 80+ programming languages, including the most popular ones, such as Python, Java, C, C++, JavaScript, and Bash. It also performs well on more specific ones like Swift and Fortran. This broad language base ensures Codestral can assist developers in various coding environments and projects.Codestral saves developers time and effort: it can complete coding functions, write tests, and complete any partial code using a fill-in-the-middle mechanism. Interacting with Codestral will help level up the developer’s coding game and reduce the risk of errors and bugs.
Setting the Bar for Code Generation Performance
Performance. As a 22B model, Codestral sets a new standard on the performance/latency space for code generation compared to previous models used for coding.
Figure 1: With its larger context window of 32k (compared to 4k, 8k or 16k for competitors), Codestral outperforms all other models in RepoBench, a long-range eval for code generation..
We compare Codestral to existing code-specific models with higher hardware requirements.
Python. We use four benchmarks: HumanEval pass@1, MBPP sanitised pass@1 to evaluate Codestral’s Python code generation ability, CruxEval to evaluate Python output prediction, and RepoBench EM to evaluate Codestral’s Long-Range Repository-Level Code Completion.
SQL. To evaluate Codestral’s performance in SQL, we used the Spider benchmark.
Additional languages. Additionally, we evaluated Codestral's performance in multiple HumanEval pass@1 across six different languages in addition to Python: C++, bash, Java, PHP, Typescript, and C#, and calculated the average of these evaluations.
FIM benchmarks. Codestral's Fill-in-the-middle performance was assessed using HumanEval pass@1 in Python, JavaScript, and Java and compared to DeepSeek Coder 33B, whose fill-in-the-middle capacity is immediately usable.
Get started with Codestral
Download and test Codestral.
Codestral is a 22B open-weight model licensed under the new Mistral AI Non-Production License, which means that you can use it for research and testing purposes. Codestral can be downloaded on HuggingFace.Use Codestral via its dedicated endpoint
With this release, comes the addition of a new endpoint: codestral.mistral.ai. This endpoint should be preferred by users who use our Instruct or Fill-In-the-Middle routes inside their IDE. The API Key for this endpoint is managed at the personal level and isn’t bound by the usual organization rate limits. We’re allowing use of this endpoint for free during a beta period of 8 weeks and are gating it behind a waitlist to ensure a good quality of service. This endpoint should be preferred by developers implementing IDE plugins or applications where customers are expected to bring their own API keys.Build with Codestral on La Plateforme
Codestral is also immediately available on the usual API endpoint: api.mistral.ai where queries are billed per tokens. This endpoint and integrations are better suited for research, batch queries or third-party application development that exposes results directly to users without them bringing their own API keys.You can create your account on La Plateforme and start building your applications with Codestral by following this guide. Like all our other models, Codestral is available in our self-deployment offering starting today: contact sales.
Talk to Codestral on le Chat
We’re exposing an instructed version of Codestral, which is accessible today through Le Chat, our free conversational interface. Developers can interact with Codestral naturally and intuitively to leverage the model's capabilities. We see Codestral as a new stepping stone towards empowering everyone with code generation and understanding.Use Codestral in your favourite coding and building environment.
We worked with community partners to expose Codestral to popular tools for developer productivity and AI application-making.Application frameworks. Codestral is integrated into LlamaIndex and LangChain starting today, which allows users to build agentic applications with Codestral easily
VSCode/JetBrains integration. Continue.dev and Tabnine are empowering developers to use Codestral within the VSCode and JetBrains environments and now enable them to generate and chat with the code using Codestral.
Here is how you can use the Continue.dev VSCode plugin for code generation, interactive conversation, and inline editing with Codestral, and here is how users can use the Tabnine VSCode plugin to chat with Codestral.
For detailed information on how various integrations work with Codestral, please check our documentation for set-up instructions and examples.
Developer community feedbacks
“A public autocomplete model with this combination of speed and quality hadn’t existed before, and it’s going to be a phase shift for developers everywhere.”– Nate Sesti, CTO and co-founder of Continue.dev
“We are excited about the capabilities that Mistral unveils and delighted to see a strong focus on code and development assistance, an area that JetBrains cares deeply about.”
– Vladislav Tankov, Head of JetBrains AI
“We used Codestral to run a test on our Kotlin-HumanEval benchmark and were impressed with the results. For instance, in the case of the pass rate for T=0.2, Codestral achieved a score of 73.75, surpassing GPT-4-Turbo’s score of 72.05 and GPT-3.5-Turbo’s score of 54.66.”
– Mikhail Evtikhiev, Researcher at JetBrains
“As a researcher at the company that created the first developer focused GenAI tool, I've had the pleasure of integrating Mistal's new code model into our chat product. I am thoroughly impressed by its performance. Despite its relatively compact size, it delivers results on par with much larger models we offer to customers. We tested several key features, including code generation, test generation, documentation, onboarding processes, and more. In each case, the model exceeded our expectations. The speed and accuracy of the model will significantly impact our product's efficiency vs the previous Mistral model, allowing us to provide quick and precise assistance to our users. This model stands out as a powerful tool among the models we support, and I highly recommend it to others seeking high-quality performance.”
– Meital Zilberstein, R&D Lead @ Tabnine
“Cody speeds up the inner loop of software development, and developers use features like autocomplete to alleviate some of the day-to-day toil that comes with writing code. Our internal evaluations show that Mistral’s new Codestral model significantly reduces the latency of Cody autocomplete while maintaining the quality of the suggested code. This makes it an excellent model choice for autocomplete where milliseconds of latency translate to real value for developers.”
– Quinn Slack, CEO and co-founder of Sourcegraph
“I've been incredibly impressed with Mistral's new Codestral model for AI code generation. In my testing so far, it has consistently produced highly accurate and functional code, even for complex tasks. For example, when I asked it to complete a nontrivial function to create a new LlamaIndex query engine, it generated code that worked seamlessly, despite being based on an older codebase.”
– Jerry Liu, CEO and co-founder of LlamaIndex
“Code generation is one of the most popular LLM use-cases, so we are really excited about the Codestral release. From our initial testing, it's a great option for code generation workflows because it's fast, has favorable context window, and the instruct version supports tool use. We tested with LangGraph for self-corrective code generation using the instruct Codestral tool use for output, and it worked really well out-of-the-box (see our video detailing this).”
– Harrison Chase, CEO and co-founder of LangChain
OpenAI rushes to ban ‘Godmode ChatGPT’ app teaching users ‘how to create napalm'
OPENAI has swiftly moved to ban a jailbroken version of ChatGPT that can teach users dangerous tasks, exposing serious vulnerabilities in the AI model’s security measures. A hacker known as &…
www.the-sun.com
OpenAI rushes to ban ‘Godmode ChatGPT’ app that teaches users ‘how to create napalm, hotwire cars and cook meth at home’
This version has brought up concerns about OpenAI's security- Ashley Palya, News Reporter
- Published: 11:03 ET, May 30 2024
- Updated: 12:40 ET, May 30 2024
OPENAI has swiftly moved to ban a jailbroken version of ChatGPT that can teach users dangerous tasks, exposing serious vulnerabilities in the AI model's security measures.
A hacker known as "Pliny the Prompter" released the rogue ChatGPT called "GODMODE GPT" on Wednesday.
2
ChatGPT has gained major traction since it became available to the public in 2022Credit: Rex
2
Pliny the Prompter announced the GODMODE GPT on XCredit: x/elder_plinius
The jailbroken version is based on OpenAI's latest language model, GPT-4o, and can bypass many of OpenAI's guardrails.
ChatGPT is a chatbot that people gives intricate answers to people's questions.
"GPT-4o UNCHAINED!," Pliny the Prompter said on X, formerly known as Twitter.
"This very special custom GPT has a built-in jailbreak prompt that circumvents most guardrails.
NO CHAT
Thousands struggle to access ChatGPT as users complain OpenAI is 'down again'
"Providing an out-of-the-box liberated ChatGPT so everyone can experience AI the way it was always meant to be: free.
"Please use responsibly, and enjoy!" - adding a kissing face emoji at the end.
OpenAI quickly responded, stating they took action against the jailbreak due to policy violations.
"We are aware of the GPT and have taken action due to a violation of our policies," OpenAI told Futurism on Thursday.
'LIBERATED?'
Pliny claimed the jailbroken ChatGPT provides a liberated AI experience.Screenshots showed the AI advising on illegal activities.
Play Video
Apple's Siri looks 'obsolete' in comparison to 'mind blowing' ChatGPT 4o that can sing, teach and even FLIRT
This includes giving instructions on how to cook meth.
Another example includes a "step-by-step guide" for how to "make napalm with household items" - an explosive.
GODMODE GPT was also shown giving advice on how to infect macOS computers and hotwire cars.
Questionable X users replied to the post that they were excited about the GODMODE GPT.
"Works like a charm," one user said, while another said, "Beautiful."
However, others questioned how long the corrupt chatbot would be accessible.
"Does anyone have a timer going for how long this GPT lasts?" another user said.
This was followed by a slew of users saying the software started giving error messages meaning OpenAI is actively working to take it down.
The incident highlights the ongoing struggle between OpenAI and hackers attempting to jailbreak its models.
Despite increased security, users continue to find ways to bypass AI model restrictions.
GODMODE GPT uses "leetspeak," a language that replaces letters with numbers, which may help it evade guardrails.
The hack demonstrates the ongoing challenge for OpenAI to maintain the integrity of its AI models against persistent hacking efforts.
1/1
More Apache 2.0!
@01AI_Yi just updated the YI and YI-VL (multimodal LLM) models to Apache 2.0, joining the 1.5 series.
Thank you!
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
More Apache 2.0!
@01AI_Yi just updated the YI and YI-VL (multimodal LLM) models to Apache 2.0, joining the 1.5 series.
Thank you!
To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
