#AI Newsletter

10 minutes with #AI or 5 stories from the world of AI [vol. 57]

11 Feb 2025
โˆ™

Each week, we bring you 5 stories that resonated the most in our internal Slack channel #AI-news. We write the newsletter using various AI tools because we're an AI company and our marketing wants to move with the times too. ๐Ÿ˜Ž

Today you're reading the 57th issue in a row.

#1

AI for $200/month? OpenAI introduces Deep Research, a new tool for researchers ๐Ÿ“Š๐Ÿ’ก.

OpenAI ๐ŸŽฏ has unveiled Deep Research, a new AI feature for Pro users at $200/month that quickly processes large amounts of data ๐Ÿ“Š. It aids research by summarizing texts, reading charts, and extracting key information from PDFs in minutes instead of hours or days ๐Ÿ•ฐ๏ธ.

The model better understands complex topics ๐Ÿง , but still struggles with real-world context and uncertainty. It can connect existing knowledge and apply methods, but it does not create new theories ๐Ÿ“– or challenge established scientific processes ๐Ÿ”ฌ.

Scientists ๐Ÿ›๏ธ discover new insights by questioning assumptions and identifying gaps in understanding ๐Ÿค”. AI, which learns only from existing data, follows established patterns instead of changing them. That makes it a useful tool, but it cannot replace scientific discoveries ๐Ÿš€. What do you think about this ๐Ÿค”?

Article on forbes.com

 

#2

Gemini 2.0 ๐Ÿค–: Google has made its most advanced AI model suite available to everyone ๐Ÿš€!

Google has released Gemini 2.0, its most advanced AI model suite, for everyone ๐Ÿค–. It includes models designed for programming, processing large amounts of data, and cost-effective AI applications ๐Ÿ’ก.

This move is part of Google's plan to develop AI agents ๐Ÿง  capable of handling complex tasks without human assistance ๐Ÿš€. Other major companies, including Meta, Amazon, Microsoft, OpenAI, and Anthropic, are also investing in this technology ๐Ÿ—๏ธ.

AI agents can automate tasks โš™๏ธ such as filling out forms, making reservations, or searching for information ๐Ÿ”. Google aims to outpace the competition with enhanced capabilities that allow its models to work better with text, images, and sound ๐ŸŽต.

Article on cnbc.com

 

#3

Mega AI Center in France ๐Ÿค–! $52 billion from French and Arab investors ๐Ÿ’ฐ๐Ÿค”?

France and the United Arab Emirates have agreed to invest up to $52 billion ๐Ÿ’ฐ in a massive AI data center ๐Ÿค– in France. This is part of France's strategy to attract AI investments ahead of the Artificial Intelligence Action summit in Paris.

The data center will have a capacity of 1 GW โšก and is funded by both French and Emirati investors, including MGX ๐Ÿ“ˆ, a UAE investment firm focused on AI. France aims to become a top destination for AI infrastructure, leveraging its nuclear and renewable ๐ŸŒ energy mix.

This investment signals that Europe doesnโ€™t want to fall behind in the AI race ๐Ÿ. More major announcements are expected at the AI summit, where political and tech leaders ๐Ÿค will discuss funding, regulations โš–๏ธ, and AIโ€™s environmental impact ๐ŸŒฑ. What do you think ๐Ÿค”?

Article on techcrunch.com

 

#4

GeoSpy ๐ŸŒ: AI tool for locating photo origins sparks privacy concerns โš ๏ธ!

A new AI breakthrough has surfaced. The AI tool GeoSpy can analyze any image ๐Ÿ–ผ๏ธ and determine where it was taken using details like buildings, plants, and streets ๐ŸŒฟ. Originally designed for law enforcement and the military ๐Ÿ•ต๏ธโ€โ™‚๏ธ, it helps track criminals, find missing persons, and solve cold cases.

For a short time, the public also had access to the tool, leading to cases of stalking and doxxing ๐Ÿ”โš ๏ธ. Although the company quickly restricted access, the damage was doneโ€”AI is making it easier than ever to uncover private information ๐Ÿ”“ from ordinary photos. Even without location tags, it can pinpoint where a picture was taken, raising serious privacy concerns.

This is yet another reminder that anything shared online can be analyzed in ways we never expected ๐Ÿคฏ. The internet is entering a new era where protecting privacy ๐Ÿ›ก๏ธ is harder than ever ๐ŸŒ.

Article on howtogeek.com

 

#5

Super Bowl 2025 ๐Ÿˆ: Google faked Geminiโ€™s output in its Super Bowl ad ๐Ÿคฏ.

And finally, a little โ€œscandalโ€ ๐Ÿคฏ. During this yearโ€™s Super Bowl, Google aired an ad showing Gemini generating a website description for a cheese shop ๐Ÿง€. The problem? The exact same text had existed since 2020โ€”three years before Gemini was even created ๐Ÿคจ. When questioned, Google claimed AI was involved, but the explanation didnโ€™t quite add up ๐Ÿค”.

This comes after criticism of the ad for incorrect data on global cheese consumption ๐Ÿ“‰. Google quietly removed the false claim but still labeled the description as โ€œAI-generatedโ€ ๐Ÿค–.

This raises an important question โ“: are tech companies showcasing AIโ€™s real capabilities, or just what they wish AI could do ๐ŸŽญ? What do you think ๐Ÿค”?

Article on theverge.com

 


Discover the full scale of Cequence's capabilities

Join the portfolio of our satisfied customers from large enterprises to medium businesses and learn how to improve your business today.

50,000 + people use Cequence to manage their contracts

DellIntelscantraxx