The Prompt is a weekly look at the fastest-growing startups, biggest breakthroughs, and commercial deals in AI. To receive it in your inbox, subscribe here.
Welcome back to The Prompt.
Today, OpenAI announced that it recently began training a new AI model that pass GPT-4 (the model that powers ChatGPT) and produce what it calls the “next level of capabilities” (although it hasn’t specified what those capabilities are.) The AI company also said it’s creating a new safety and security committee assess the risks posed by the technology and formulate an initial set of recommendations to mitigate them over the next 90 days. The committee will be led by members of the company’s board of directors, including Bret Taylor, Adam D’Angelo, Nicole Seligman and CEO Sam Altman. OpenAI said the committee would consult external experts such as Rob Joyce, former director of cybersecurity at the National Security Agency, and John Carlin, former Justice Department official.
Now let’s move on to the headlines.
ETHICS + LAW
In a TikTok posted earlier this month, Cassey Ho said Begoing face-swapped in a video where she is seen modeling her own clothes.
Cassey Ho
At a fashion designer the body was tampered with using AI tools to selling rip-offs of their own products on platforms like Amazon and Walmart. Forbes spoke with Cassey Ho, CEO and founder of fashion brand Popflex and creator of viral clothing item, the “pirouette cards” which, according to her, is counterfeit. Ho said she found the deepfake”incredibly violent.” Photos of his brand’s models and clients were also stolen and posted on Amazon ads to sell the counterfeits, she said.
MARKET MOVEMENTS
Chipmaker Nvidia had a tremendous week on Wall Street. During an earnings conference call on Wednesday, the AI giant announced that its profits increased by 630% and revenue increased by approximately 270% this time last year, making it from Nvidia the most profitable and highest sales quarter ever. The company’s shares hit a record high on Tuesday and peaked above $1,000 per share for the first time. (The company’s shares closed at $1,064 on Friday.) The stock’s rise added more than $10 billion to CEO Jensen Huang’s net worth, putting him in the top 20 of the company’s stocks. Forbes“Ranking of the richest people in the world.
AI HUMANS
Generative AI is increasingly making its way into the world of real estate, whether it is design fantastic houses, write descriptions announcements or scan and process documents for home loan approvals. Now, a new AI assistant called Sidekick aims to empower thousands of real estate agents research properties, compare their prices and create ad descriptions from uploaded photos. AI is a powerful tool for real estate agents because they’re “salespeople, they’re not experts in data analysis, which is what they need to do,” said Michael Martin, co-founder and CEO of Sidekick. Currently focused on the Miami and San Francisco areas, Sidekick is built on GPT-4 and trained on a huge real estate database called MLS.
AI DEALS OF THE WEEK
Defense AI company Helsing is in talks to raise $400 million for a valuation of 4 billion dollarssources said Forbes. Led by General Catalyst, the round would bring the company’s total investment to $750 million. The Munich-based company has won contracts to provide AI software to help German fighter jet pilots and develop AI capabilities for drones in Ukraine.
Based in Germany AI translation startup DeepL raised $300 million for a valuation of $2 billion focus on business growth. The company’s proprietary translation model is used by more than 100,000 companies to translate documents in over 30 languages and recently launched an AI-powered business writing tool.
The French AI startup H announced that it had raised $220 million in seed fundingo create AI agents for business applications. Founded by former Stanford and Google DeepMind researchers in 2023, the company is supported by Eric Schmidt, Accel and Amazon, among others.
DEEP DIVE
As Google rolls out its spectacularly bad AI search results, competitors like You.com and Duck Duck Go say its monopoly remains “entrenched.”
Getty Images
Days after Google integrated AI insights into its flagship search engine, the platform used by billions of people around the world produced a series of misleading responses. Google told people they could to use stick to make sure the cheese sticks to the pizzathat they could watch the sun for up to 30 minutes and that geologists suggest eat one rock per day. It also produced other inaccuracies, such as the claim that former President Barack Obama is Muslim and failing to find existing information, such as the names of countries in Africa beginning with K. After that As stories about these errors went viral on social media, Google removed some of these errors. these answers. Part of the problem? Google’s AI search insights appeared to come from unverified sources such as Reddit posts and processed satirical sites such as The onion as legitimate sources of information.
Representatives of alternative search platforms like DuckDuck Go and You.com said that even if Google’s AI generates inaccurate information, the tech giant maintains a strong monopoly in researchlargely because it is the default search engine on popular devices and browsers. “Even if the results are worse or people don’t like the product, can they change?” DuckDuckGo spokesperson Kamyl Bazbaz said in an email highlighting the U.S. Department of Justice’s antitrust lawsuit against Google, which alleges the Search giant pays Apple more than $20 billion be the default search engine for Safari.
You.com CEO Richard Socher also expressed frustration with Google’s apparent stranglehold on search, saying that while he has seen a 5 to 10 percent increase in usage of its search engine, AI-powered conversational search after Google rolled out its AI-generated answers in search, it’s not enough to significantly increase market share. “I think there’s an undercurrent that people are just fed up with Google” said Socher Forbes. “But it seems impossible. The monopoly is very rooted.”
YOU HAVE AN INDEX
1 of 6
Answers generated by AI models for legal queries are incorrect, according to a new Stanford study.
75%
Answers to questions about court decisions created by large general-purpose language models like OpenAI’s GPT 3.5, Google’s PaLM 2 and Meta’s Llama 2 are also incorrect, according to a previous study.
QUIZ
Elon Musk is involved in an online feud with this AI leader.
- Demis Hassabis
- Sam Altman
- Yann LeCun
- Clement Delangue
Check if you have it here.
MODEL BEHAVIOR
A new app called Style DNA uses AI to browse a person’s existing wardrobe to make style and outfit recommendations. The UK-based company claims its app has over 3 million downloads and 300,000 active users. It works by scanning a person’s selfie to assess their skin tone and features and suggest a color palette for their outfits. Co-founded by Elena Volkova in 2019, the personal styling company aims to use AI to combat “nothing to wear” dilemma.