šŸšØExperts Sound the Alarm on AI

In todayā€™s email:

šŸšØ A new AI report is raising serious concernsā€”from terrorism to job losses and AI malfunctions. Experts are sounding the alarm, urging policymakers to act before things get out of hand.

šŸ‘©ā€šŸ’» AI needs more women! With biases already baked in, it's crucial that women get involved in shaping AI's future. Hereā€™s why it's not just important but necessary.

šŸ’” Missed some AI breakthroughs? From synthetic data to a new AI model that predicts without real-world data, hereā€™s the latest on whatā€™s happening behind the scenes.

Intrigued? Keep scrolling!

šŸš© AI Report Raises Red Flags

Designed by Freepik

A new report about the risks of general-purpose AI is out, and itā€™s got experts worried. Theyā€™re saying that while AI has come a long way (from writing bad paragraphs to writing whole computer programs and creating fake images), itā€™s not all sunshine and rainbows. The AI weā€™re talking about could lead to some serious problems, like:

  • Terrorism: AI could be used to help create chemical/biological weapons.

  • Job losses: It might replace workers in various industries (though some think it'll create jobs too).

  • Malfunctions: Systems could go haywire, either by messing up on their own or because humans aren't paying attention.

Even though AI is still in the ā€œletā€™s figure it outā€ phase, the risks are growing as it gets more capable. Experts can't even agree on when itā€™ll surpass humans in some areas, but they do know the potential outcomes range from ā€œsuper helpfulā€ to ā€œtotal disaster.ā€

The report was put together by 100 experts and backed by 30 countries (yes, even the US and China agreed on something), and itā€™s being released ahead of the big AI summit in Paris. The goal is to get policymakers thinking ahead about these risks and figuring out how to keep things from spiraling out of control.

šŸ‘© AI Needs More Women

Designed by Freepik

AI could either be the best thing to happen for gender equality or a total disaster, depending on whoā€™s building it. Right now, women make up less than a third of AI professionals, only 18% of researchers, and are way less likely than men to use AI tools at work. Thatā€™s a problem. But guess what? You donā€™t need a PhD in machine learning to start using AIā€”you just need curiosity and a smartphone.

Why should women care?

  • AI could automate annoying tasks (hello, fewer emails and spreadsheets).

  • It can boost confidenceā€”helping with public speaking, salary negotiations, and rewriting emails that are way too apologetic.

  • Itā€™s like having a mentor in your pocket, offering career advice without judgment.

The bad news?

  • AI is already biasedā€”favoring men for jobs, sexualizing women in images, and reinforcing stereotypes.

  • Deepfakes and privacy issues disproportionately harm women.

The solution?

Women need to get involved, start using AI, and help shape its future. Otherwise, old biases will be coded into the tech that will define our lives. So, start smallā€”ask ChatGPT for help, share AI tips with friends, and donā€™t let men have all the fun. The future is being built nowā€”letā€™s make sure women are part of it!

šŸ“ˆ AI Breakthroughs You Missed

Designed by Freepik

2025 is already shaping up to be a wild ride in AI. Trump just threw $500 billion at an AI initiative called the Stargate Project. But just a day later, a Chinese AI company, DeepSeek, basically said, "Hold my beer" and dropped DeepSeek-R1, a large language model that works just as well as OpenAIā€™s top models but at a fraction of the cost. Naturally, tech stocks had a meltdown.

Now, while all the flashy AI stuff gets the spotlight, some lesser-known but game-changing research is happening behind the scenes. One such study, published in Nature, introduces TabPFN, an AI model that can analyze spreadsheets and make highly accurate predictions without using real-world data. Instead, it learns from 100 million synthetic datasetsā€”randomly generated but statistically legit. This could be huge for fields like finance, medicine, and genomics.

Why this matters:

  • AI usually needs tons of real data, but what if there isnā€™t enough?

  • Synthetic data could fill the gapā€”if itā€™s done right.

  • Thereā€™s still a risk of AI "hallucinating" bad results, so replication is key.

Of course, AI safety remains a concern, especially since Trump just scrapped an executive order focused on AI trust and security. So, while weā€™re all gawking at AIā€™s latest tricks, letā€™s not forget the researchers trying to actually understand how and why AI works.

Other cool AI stuff that is trending right now šŸ”„šŸ”„

šŸ”“ An ex-OpenAI safety researcher just hit the panic button, calling AIā€™s breakneck speed ā€˜terrifying.ā€™ - Read more

šŸ’° Abu Dhabi is dropping $13 billion to become the worldā€™s first fully AI-powered government by 2027. - Read more

šŸ„¼ OpenAIā€™s teaming up with U.S. national labs to bring AI into nuclear weapons research. - Read more

šŸ—ŗļø Generative AI is not just reshaping industriesā€”itā€™s sparking a global tech tug-of-war, where how nations play their cards could determine whoā€™s top dog on the digital world stage. - Read more

šŸ¤– DeepSeek, the Chinese AI that's making waves, is now under investigation in Europe for potentially putting millions of personal data at risk. - Read more

What Are Your Thoughts Of Today's Email?

Login or Subscribe to participate in polls.