The AI Race: Are we pushing the limits or just chasing hype?

As artificial intelligence has become a hot topic, the biggest companies have entered a technology race. Giants are competing to release new and improved AI models, offering users more features and promising even better results.

But is that really true? How far can AI models be developed, and is it even necessary? In today's edition of the newsletter, we’ll take a closer look at what the future of AI models might bring, both in the short and long term.

?? Scientists investigate: Will AI take over the world?

As AI becomes more accessible, concerns and doubts have emerged. People are asking, "Will it replace us?" According to researchers from the University of Bath and the Technical University of Darmstadt, that's not possible.

Their research shows that while large language models (LLMs) can follow instructions and are good at using language, But they cannot develop complex reasoning skills without clear guidance. They are unable to learn new skills on their own and do not pose an existential threat to humanity.

How did the scientists reach this conclusion, and what else is worth knowing on the topic? Find out in the article: https://www.roboticsandautomationmagazine.co.uk/news/ai/ai-poses-no-existential-threat-to-humanity-university-study-finds.html

?? AI crisis: LLM systems may enter a vicious cycle

Tech companies continue to release new models, but AI experts are raising concerns — “Model collapse” could happen. Current AI models need high-quality human data to work properly. If they are trained only on AI-generated data, it leads to a drop in quality and diversity.

With the growing availability of AI-generated content, we could enter a vicious cycle. New models will use content created by older models, which will cause the quality of the outputs to decrease.

Of course, the situation isn’t set in stone yet, but one thing is clear: for AI to stay effective, humans cannot remain passive.

Read the article and learn what tech companies are already doing to prevent this crisis: https://www.fastcompany.com/91175349/generative-ai-model-collapse-tech-companies

?? Tech news: Microsoft releases 3 new AI models

Despite the growing challenges with LLMs, the tech giant isn’t slowing down. Microsoft has released three new models from the Phi 3.5 series, aimed at advancing artificial intelligence even further. Each model is designed for specific tasks, allowing better adaptation to different developer and application needs.

All models are available under an open MIT license, allowing free use, modification, and commercial applications without restrictions.

Check out the details of these models and find out what they can do: https://venturebeat.com/ai/microsoft-releases-powerful-new-phi-3-5-models-beating-google-openai-and-more/

The world of artificial intelligence is constantly changing, and the future of AI models brings both exciting innovations and serious challenges. New models offer incredible possibilities — from advanced language and image processing to the automation of more complex tasks. However, there are also questions about how far we can push the development of this technology and whether the rapid release of new models is truly beneficial.

Does this technological race meet users' needs? Or could this fast development lead to problems, like a decrease in model quality and diversity?

Leave a comment and share your thoughts! ??

? Don’t miss - Europe’s first TED CONFERENCE on AI! ?

If you're interested in the future of artificial intelligence and want to learn more about how AI is shaping our world, don't miss the upcoming TEDxAI Vienna event. It's a unique opportunity to hear inspiring talks from AI experts and innovators sharing their insights on the technology of tomorrow.

Check out the details and register here: https://tedai-vienna.ted.com/?gad_source=1&gclid=CjwKCAjwoJa2BhBPEiwA0l0ImMUXnaqce9fTHayWnm5f1541Oz4V-1w_VE_BSHXd_WDIdTXhPMDAdhoCSeMQAvD_BwE