September 23, 2025

news  
Meta Invests Billions, Google DolphinGemma, ChatGPT Picks Stocks

The world of artificial intelligence is rapidly expanding, touching everything from financial markets to interspecies communication. In finance, AI trading tools promise efficiency but raise concerns about market chaos, herd behavior, and hidden costs for individual investors, drawing parallels to the dot-com bubble. Experts caution that technology alone doesn't guarantee profits, emphasizing the need for human judgment and risk management. Meanwhile, Meta CEO Mark Zuckerberg is heavily investing billions in AI, viewing the risk of falling behind in the race for superintelligence as greater than the possibility of an AI bubble. He believes that while some companies may overspend, hesitating in AI development could leave Meta out of position. On the research front, scientists are using AI, including Google's DolphinGemma model, to attempt communication with dolphins, while MIT researchers are leveraging AI for mathematical discovery and automated theorem proving. In education, UNC Charlotte is launching new AI and construction engineering degrees. The potential economic impact of AI is significant, with the World Trade Organization predicting a nearly 40% boost in global trade by 2040, though it warns of widening economic divides between developed and developing nations. In the realm of entertainment and media, an AI series called 'Whispers' is exploring interactive storytelling, and an independent baseball team experimented with AI game management, sparking fan backlash. The rapid adoption of AI also presents security challenges, with many IT leaders lacking confidence in their defenses against AI-driven cyberattacks. Even consumer applications are seeing AI integration, as demonstrated by ChatGPT's GPT-5 model picking aggressive stocks for a small investment.

Key Takeaways

  • AI trading tools are raising concerns about market volatility and potential chaos due to herd behavior, with experts drawing parallels to the dot-com bubble.
  • Meta CEO Mark Zuckerberg is prioritizing significant AI investment, believing the risk of falling behind in AI development outweighs the possibility of an AI bubble.
  • The World Trade Organization forecasts AI could increase global trade by nearly 40% by 2040 but warns of widening economic divides between nations.
  • Scientists are using AI models, like Google's DolphinGemma, to study dolphin communication, aiming for potential two-way conversations.
  • MIT researchers are receiving grants to use AI for accelerating mathematical discovery and improving automated theorem proving.
  • UNC Charlotte is introducing new Bachelor of Science degrees in Artificial Intelligence and Construction Engineering starting in fall 2026.
  • A significant portion of IT leaders lack confidence in their organization's security defenses against AI-driven cyberattacks, despite acknowledging increased cybercrime due to AI.
  • ChatGPT's GPT-5 model demonstrated aggressive stock picking for a small investment, selecting high-growth stocks like Palantir.
  • An AI series titled 'Whispers' is debuting interactive storytelling formats, exploring new narrative possibilities with AI.
  • An independent baseball team's experiment with AI game management led to fan backlash, highlighting concerns about AI replacing human roles and corporate priorities.

AI trading tools create hidden costs for everyday investors

AI is changing financial trading, offering speed and efficiency that often benefit big investors more than individuals. Everyday investors face challenges like not having the same real-time data or fast trade execution. Many AI trading systems are also complex and lack transparency, making it hard for regular people to understand the risks. While some AI tools are available, the most powerful ones can be expensive. To navigate this, investors should focus on education, long-term strategies, and transparent platforms.

AI investment tools could cause market chaos, experts warn

Wall Street firms are creating AI investment tools they say will help ordinary Americans get financial advice. However, experts worry that if everyone uses the same AI strategies, it could lead to herd behavior and increase market ups and downs. These tools might also reduce the need for human advisors who understand individual needs. The 'black box' nature of some AI also raises concerns about transparency and blindly following advice. A balanced approach using AI alongside human judgment is seen as crucial.

AI trading hype echoes dot-com bubble dangers

The excitement around AI investment tools today is similar to the speculation before the dot-com crash. While AI offers advanced analysis and speed, investors may still make emotional decisions leading to losses. Retail investors might rely too much on AI without understanding its limits, repeating mistakes from the past. Experts warn that technology alone doesn't guarantee profits, and sound investment strategies with risk management are still essential. The danger lies in seeing advanced tech as a guaranteed path to wealth.

AI trading hype mirrors dot-com bubble, warns industry

The current excitement around AI trading tools is drawing comparisons to the dot-com bubble of the late 1990s. While today's AI is backed by corporate earnings, the rapid rise in AI stocks and the speed of trading raise concerns about overvaluation and market mania. Experts warn that AI can introduce new risks like algorithmic biases and flash crashes. Regulatory gaps and the dominance of big tech in AI add to potential dangers. History suggests that unchecked hype can lead to painful corrections, urging caution and due diligence.

Meta CEO Zuckerberg invests heavily in AI despite bubble fears

Meta CEO Mark Zuckerberg believes the risk of falling behind in AI is greater than the possibility of an AI bubble. He plans to invest billions in artificial intelligence, even acknowledging that companies might overspend and fail. Zuckerberg stated that hesitating in AI development could leave Meta 'out of position' if another company achieves superintelligence first. Unlike some AI startups, Meta's ad business provides a stable foundation, reducing the risk of business failure from AI investments.

Zuckerberg sees AI bubble risk but prioritizes AI investment

Meta CEO Mark Zuckerberg acknowledges that an AI bubble is possible, comparing it to past historical bubbles. However, he argues that the greater risk is not investing heavily enough in AI, which could cause Meta to miss out on advancements in superintelligence. Zuckerberg explained that while misspending billions would be unfortunate, the risk of being left behind is higher. He noted that past infrastructure buildouts, like railroads and the dot-com era, sometimes led to bubbles but ultimately created valuable networks.

AI could boost global trade but may widen economic divides

A new World Trade Organization report suggests artificial intelligence could increase global trade by nearly 40% by 2040, boosting economies. However, the report also warns that AI could worsen economic inequality between nations. Developed countries with advanced technology may benefit more, leaving developing nations behind. The WTO recommends policies to ensure AI adoption is inclusive, including investing in digital infrastructure and training. International cooperation is needed to share AI's benefits widely and prevent a growing digital divide.

ChatGPT's GPT-5 picks surprising stocks for $500 investment

A journalist gave ChatGPT, using its GPT-5 model, $500 to invest in stocks for six months, expecting basic advice. Instead, the AI chose five aggressive, high-growth stocks: Palantir (PLTR), AppLovin (APP), Agios Pharmaceuticals (AGIO), Hut 8 Corp. (HUT), and MicroStrategy Inc. (MSTR). The AI explained its choices were based on extensive research and aimed to maximize returns. This experiment contrasts with earlier AI models that gave more generic recommendations.

UNC Charlotte launches new AI and construction engineering degrees

UNC Charlotte will offer three new degrees starting in fall 2026: a Bachelor of Science in Artificial Intelligence, a Master of Science in Artificial Intelligence, and a Bachelor of Science in Construction Engineering. These programs were approved by the UNC Board of Governors. The AI degrees will position the university among a select group nationally, while the construction engineering degree aims to address a critical workforce shortage in the industry. A new online Bachelor of Arts in Computer Science will also be available next fall.

AI helps researchers attempt to talk to dolphins

Scientists are using Artificial Intelligence (AI) to try and understand dolphin communication. A Google and Georgia Tech project trained an AI language model called DolphinGemma using 40 years of dolphin sound data. The AI can now produce sounds similar to dolphin clicks and whistles. Researchers are developing a device called CHAT to test if these AI sounds can be used to communicate with dolphins, potentially leading to a two-way conversation. They hope this research will highlight dolphin intelligence and change how people view these animals.

IT leaders lack confidence in AI security defenses

A recent report shows that only 31% of IT leaders feel confident in handling AI-driven cyberattacks, despite 61% noting increased cybercrime due to AI. Many IT leaders are concerned about employees using public AI tools and the risks associated with organizational AI agents. Confidence in mitigating these risks is low, with only about a third feeling their vulnerability analysis is adequate for AI threats. Accelerated AI adoption may be creating blind spots in security defenses.

Oakland Ballers AI game management sparks fan backlash

The Oakland Ballers, an independent baseball team, let an AI manage a game, using ChatGPT trained on baseball data. The AI made decisions similar to the human manager, Aaron Miles, who saw it as a tool to optimize strategy. However, the experiment angered some fans who felt it represented corporate greed and a focus on the AI race over tested products, similar to other sports franchises leaving Oakland. The team's management acknowledged the fan reaction and does not plan to repeat the AI experiment.

MIT researchers win grants for AI-driven math discovery

MIT researchers David Roe and Andrew Sutherland are among the first recipients of the AI for Math grant, aiming to speed up mathematical discovery. They will use the funding to connect the L-Functions and Modular Forms Database (LMFDB) with the Lean4 mathematics library (mathlib). This project will improve automated theorem proving by making a large database of unformalized mathematical knowledge accessible to AI systems. The goal is to help mathematicians and AI agents find new theorems and proofs more efficiently.

AI series 'Whispers' debuts interactive storytelling at Busan market

Pickford's new AI series 'Whispers' was showcased at the Busan Asian Contents & Film Market, featuring an interactive teaser. Producer Bernie Su created the series to explore new narrative formats only possible with AI. Audiences used QR codes to interact through a chat room, influencing the story in real time. Voice actors discussed their experiences with AI voice synthesis, noting how the technology altered their recorded dialogue. The project emphasized the importance of consent and transparency when using actors' voices with AI.

Sources

AI trading AI investment AI finance AI risks AI transparency AI market chaos AI bubble dot-com bubble AI stocks AI hype AI strategy AI regulation AI adoption AI investment strategy AI business AI development AI superintelligence AI infrastructure AI global trade AI economic divides AI inequality AI policies AI digital infrastructure AI training AI cooperation AI digital divide ChatGPT GPT-5 AI stock picking AI education AI degrees AI construction engineering AI workforce AI computer science AI dolphin communication AI language model AI sound data AI cyberattacks AI security AI cybercrime AI vulnerability AI game management AI baseball AI math discovery AI theorem proving AI mathematics AI storytelling AI narrative formats AI voice synthesis AI consent