AI-Generated Images Of Hurricane Helene Are Flooding Social Media Gaylord Contreras | usagoldmines.com

The Immediate is a weekly rundown of AI’s buzziest startups, largest breakthroughs, and enterprise offers. To get it in your inbox, subscribe here.

ASHEVILLE, NORTH CAROLINA: Heavy rains from hurricane Helene prompted document flooding and harm on September 28, 2024 in Asheville, North Carolina. (Picture by Melissa Sue Gerrits/Getty Photographs)

Getty Photographs

Welcome again to The Immediate,

As Hurricane Helene wreaks havoc in North Carolina and Florida, AI-generated photos of victims and first responders are spreading on platforms like Instagram, Fb and X, Forbes reported. An AI-generated picture of a younger woman stranded in flood waters holding a drenched pet in her hand has greater than 1,000,000 views on X, Kevin Guo, CEO of content material moderation platform Hive confirmed with me. The picture is getting used to spread misinformation in regards to the federal authorities’s response to the catastrophe.

Different AI-generated photos of Hurricane Helene’s aftermath, equivalent to a person wading by means of water whereas holding a canine, regulation enforcement officers offering reduction and Donald Trump in a life jacket trawling by means of muddy water, are additionally circulating on social media. Experts say such faux photos may harm humanitarian efforts and make it troublesome for victims to seek out assist and reliable info.

Now let’s get into the headlines.

ETHICS + LAW

AI unicorn Character AI briefly hosted a chatbot with the likeness of a teenage woman who was brutally murdered 18 years in the past, Forbes reported. The woman’s father, Drew Crecente, received a Google Alert on Wednesday morning that somebody had used his daughter’s yearbook photograph and her title to create the bot. “It shocks the conscience, and it is unacceptable conduct,” he mentioned. The corporate eliminated the chatbot from its platform for violating its impersonation insurance policies.

This comes after the once-buzzy startup’s CEO Noam Shazeer was employed by Google in an enormous $2.7 billion deal and its interim chief introduced that it plans to cease coaching innovative AI fashions and as a substitute deal with its shopper product.

PEAK PERFORMANCE

Two pc scientists, Geoffrey Hinton and John Hopfield, have been awarded the 2024 Nobel Prize in Physics for inventing methods that enable computer systems to know info in a method much like the human mind — educating programs to ingest huge quantities of information and determine patterns. Hilton, often called one of many godfathers of AI, and Hopfield used ideas from physics to coach synthetic neural networks, laying a basis for the event of machine studying programs like ChatGPT.

AI DEAL OF THE WEEK

Final week, OpenAI closed a $6.6 billion funding round at $157 billion valuation — the largest enterprise capital funding spherical of all time. Thrive Capital led the spherical with participation from giants like Microsoft, SoftBank and Nvidia amongst others, in response to The Wall Street Journal. With 250 million individuals utilizing its star product, ChatGPT, the AI big plans to make use of the contemporary funding to extend its compute capability and launch extra merchandise.

Additionally notable: Braintrust, which presents software program that evaluates how an AI product performs, has raised $36 million at a $150 million valuation, Forbes reported. Early adopters of AI like Notion and Figma use Braintrust to find out how effectively its AI integrations are responding to prompts.

DEEP DIVE

On a heat night time in San Francisco earlier this week, Elon Musk stood on high of a small spherical espresso desk, bathed in purple gentle. Flocks of AI engineers and researchers gathered round him, some sitting on a close-by stairwell, because the billionaire pitched them on becoming a member of his fledgling AI firm, xAI.

On the occasion, a recruitment celebration for potential staff, Musk’s major promoting level was velocity — touting the corporate’s tempo in creating merchandise, and creating AI instruments in a quick and nimble atmosphere, evaluating xAI to an SR-71 jet. “No SR-71 Blackbird was ever shot down and it solely had 1 technique: to speed up,” Musk mentioned, according to one attendee.

The situation was apt: The Pioneer Constructing within the metropolis’s Mission district, a 122-year-old former truck manufacturing unit, and most lately, headquarters of OpenAI, the juggernaut maker of ChatGPT led by cofounder Sam Altman. xAI had simply moved into the constructing, Musk mentioned, attendees instructed Forbes.

Musk has not been shy about his beef with OpenAI. He cofounded the corporate with Altman and others in 2015, earlier than leaving three years later after an alleged inner energy battle. Since then, Musk has sued the corporate twice, accusing the corporate of abandoning its mission to construct synthetic intelligence to profit humanity. (His newest swimsuit was in August, after withdrawing the same one in June.)

“He actually needs to be the antithesis to OpenAI,” one attendee, Marvin von Hagen, instructed Forbes when requested about major takeaways from the night. “They actually need to go on and say, ‘Okay, we’re the great guys now.’”

Musk’s remarks — in addition to the workplace location — illustrate the looming function OpenAI performs for xAI. Musk held the occasion on the identical day as OpenAI’s annual Dev Day, which some have imagined to be greater than a coincidence. On the occasion, Musk joked about constructing “BasedGPT.” xAI cleared out OpenAI’s furnishings from the workplace the day earlier than the occasion, however not less than one remnant nonetheless stays: A printed-out photograph cutout of OpenAI president Greg Brockman is hanging on the ceiling of one of many rooms, two attendees mentioned. (Brockman is at present on go away from OpenAI till the top of the 12 months.) Von Hagen estimated that he noticed about 10 present OpenAI staff in attendance.

Learn the total story on Forbes.

WEEKLY DEMO

ChatGPT and different chatbots in its class are vulnerable to regurgitate sure constructions of phrases in its responses. “At the moment’s quick paced world” and “play a big function in shaping” are among the many high 50 overused phrases by giant language fashions, in response to an evaluation by AI content material detection platform, GPTZero, Forbes reported. AI-generated textual content is especially troublesome to catch and such repetitive phrases are a technique of recognizing machine-written supplies, as I reported with my colleague Alex Levine at first of this 12 months.

MODEL BEHAVIOR

Would you desire a chatbot to be your wingman? Relationship app Grindr bets you’ll. The app, with a consumer base of largely homosexual and bisexual males, is testing an AI chatbot that may display screen prospects to suggest relationship companions, arrange dates and hold a observe of your conversations, the Wall Street Journal reported. On platforms like Tinder, some have already used chatbots up to now for them, Forbes reported again in February. In a single occasion, an individual constructed an auto-swiper and later used ChatGPT to speak with a number of girls, one in every of whom he ultimately married.