- Neuralletter
- Posts
- šØExperts Sound the Alarm on AI
šØExperts Sound the Alarm on AI

In todayās email:
šØ A new AI report is raising serious concernsāfrom terrorism to job losses and AI malfunctions. Experts are sounding the alarm, urging policymakers to act before things get out of hand.
š©āš» AI needs more women! With biases already baked in, it's crucial that women get involved in shaping AI's future. Hereās why it's not just important but necessary.
š” Missed some AI breakthroughs? From synthetic data to a new AI model that predicts without real-world data, hereās the latest on whatās happening behind the scenes.
Intrigued? Keep scrolling!
š© AI Report Raises Red Flags

Designed by Freepik
A new report about the risks of general-purpose AI is out, and itās got experts worried. Theyāre saying that while AI has come a long way (from writing bad paragraphs to writing whole computer programs and creating fake images), itās not all sunshine and rainbows. The AI weāre 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 āletās figure it outā phase, the risks are growing as it gets more capable. Experts can't even agree on when itāll surpass humans in some areas, but they do know the potential outcomes range from āsuper helpfulā to ātotal 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 itās 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.
š© 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 whoās 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. Thatās a problem. But guess what? You donāt need a PhD in machine learning to start using AIāyou just need curiosity and a smartphone.
Why should women care?
AI could automate annoying tasks (hello, fewer emails and spreadsheets).
It can boost confidenceāhelping with public speaking, salary negotiations, and rewriting emails that are way too apologetic.
Itās like having a mentor in your pocket, offering career advice without judgment.
The bad news?
AI is already biasedāfavoring 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 smallāask ChatGPT for help, share AI tips with friends, and donāt let men have all the fun. The future is being built nowāletās make sure women are part of it!
š 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 OpenAIās 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 datasetsārandomly 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 isnāt enough?
Synthetic data could fill the gapāif itās done right.
Thereās 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 weāre all gawking at AIās latest tricks, letās not forget the researchers trying to actually understand how and why AI works.
Other cool AI stuff that is trending right now š„š„
š“ An ex-OpenAI safety researcher just hit the panic button, calling AIās breakneck speed āterrifying.ā - Read more
š° Abu Dhabi is dropping $13 billion to become the worldās first fully AI-powered government by 2027. - Read more
š„¼ OpenAIās teaming up with U.S. national labs to bring AI into nuclear weapons research. - Read more
šŗļø Generative AI is not just reshaping industriesāitās sparking a global tech tug-of-war, where how nations play their cards could determine whoās top dog on the digital world stage. - Read more
š¤ 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? |