European Union is making moves on regulating AI

- Simplified AI -
Your gateway to discovering the world of Artificial Intelligence!

I just ordered the final Christmas presents for my family, so fingers crossed they show up within the next couple of days. I saw a recent poll that says 77% of Americans have a favorable opinion of the US post office. How…? No way that can be accurate. If that was true, UPS and Fedex probably wouldn't exist.

No newsletter next week, I'll be baking and eating Christmas cookies all day and spending time with family. Look out for our next newsletter on New Year's Eve for a look ahead of what's to come in 2024 for Simplified AI. Can't wait to talk to you all then!

A sneak peek of what’s below:

• What is Machine Learning?

• The EU looks to regulate AI

• AI is helping to put out wildfires

• Visualize the future of your favorite city!

What is Machine Learning?

Turning AI topics into a walk in the park, not a mind-boggling maze

Today we're exploring another subfield of AI that involves the development of algorithms and computer models: Machine Learning (ML)

Machine learning isn't exactly what it may seem like – it's not about big machines attending a classroom to improve their SAT scores. Instead, it allows computer models to enhance their performance by learning from data. ML enables machines to identify patterns, make predictions, and continuously refine their understanding and decision-making based on the information they receive. The more attempts they undergo and the more data they bring in, the better the algorithm becomes.

So, while you won't find a massive manufacturing tool trying to learn cursive handwriting, AI that has undergone machine learning can be applied to such tools to enhance their efficiency and precision, ultimately leading to improved and faster results.

Similar to how humans learn from their mistakes, AI can do the same, but on a much larger and faster scale. Can you stay awake 24/7 for an entire year to learn about a single topic? I'd wager my 2024 salary that you can't, but I wouldn't bet against an AI program running for that duration. Depending on its purpose, an AI tool might engage in open, unsupervised learning as we discussed in a previous newsletter, where it can learn about multiple topics simultaneously, or it might focus on a singular task, such as mastering the game of chess.

A computer program can be trained to understand the basic rules of chess, starting at a beginner level. After playing thousands of simulated games, it begins to acquire more sophisticated strategies - making the best opening moves, sacrificing pieces strategically, and ultimately achieving checkmate. We've already reached a point where even the best human chess players in the world can no longer defeat a computer because of its machine learning capabilities.

Chess is just one example of how ML can work. Here are some other examples of machine learning impacts daily life:

• Voice Assistants: better understand your voice commands over time

• Autocorrect: better understand your texting habits to predict your next word

• Google maps: better understand traffic patterns during various times of day to reroute you and save time on your commute

• Fraud detection: better understand your purchase history to alert you and your bank of potential fraudulent purchases

• Email Filters: better understand which emails you engage with and which emails should be sent to spam

Why is this important?

Machine learning is fundamental to artificial intelligence, forming the foundation for improving AI's performance and expanding the use cases with increased confidence in its output.

It elevates decision-making processes, granting individuals and organizations the ability to make informed and precise choices. ML fuels efficiency by automating tasks, and its predictive capabilities and personalization attributes enable tailored experiences.

AI News Flash

Catch the latest AI news that is making all of the headlines this week!

AI in the EU

The European Union has notoriously been tougher on tech companies than the US. They hand out more severe fines for penalties, they have stronger data protection laws, and the EU has been more aggressive in antitrust actions against the tech giants like Google and Microsoft.

So it doesn't come as a surprise that the EU is looking to tackle the next big thing in tech - regulation for AI.

Recently, a provisional deal was reached in the EU regarding AI regulation that will focus on key topics like prohibiting companies from using AI for facial recognition and social scores, limiting the ways employers use AI to track employees, and implementing large fines ranging from 7.5-30 million euros for violations.

The deal still has to be fully accepted by the governing body, but if it happens, it will be the world's first set of comprehensive rules on how AI should be governed. It's expected to be passed (albeit, probably with some minor changes) by the end of this year, but any laws likely wouldn't go into effect until 2025 at the earliest.

Why does this matter?

From the onset of AI, there have been major concerns about safety and privacy. When OpenAI released ChatGPT just over a year ago, it opened the floodgates for AI companies and innovation seems to be moving much faster than experts initially expected, which means there is a large need for some ground rules to protect people.

Large companies can misuse data, have breaches of data, and potentially engage in anti-competitive activities that block new, smaller entrants into the market. And Single actors can create scams and misinformation, leading to individual harm.

This set of rules proposed by the EU could serve as a benchmark for other countries to follow.

Other Stories We're Following:

• In a big blow to autonomous vehicles, Tesla had to "recall" almost all of their cars this week

• OpenAI (ChatGPT) signed a massive deal to get trained on the news so it can reduce hallucinations about current events

AI Company Spotlight

Explore trailblazing companies that are shaping our digital future

Pano

It wasn't that long ago that everyone in the northeast was getting alerts to stay indoors because of the poor air quality due to massive, raging wildfires - something that has historically been more normal to see on the west coast. It was eerie to walk outside and see the yellow-orange haze when you're used to seeing blue skies.

There were ~54k wildfires in North America this year, which is equal to the 10-yr average, but there's a massive difference in the number of acres burned - 2.6m acres compared to an average of 8.8m acres over 10-yrs.

While it's great that the number is decreasing, Pano is a company looking to help firefighters reduce that number even further.

By using cameras that sit atop mountains, combined with AI technology, Pano detects and pinpoints new fires and can alert fire professionals within minutes. They help firefighters get to the scene faster, with the right equipment and latest information, so they can stop a new wildfire in its tracks.

The company monitors more than 6 million acres of land currently and have partnered with T-Mobile to expand their coverage and make manned watchtowers an archaic practice.

Why does this matter?

This is a little bit of an upgrade on Smokey the Bear asking you to stay vigilant while you're hiking in the woods.

Employing this technology over vast land areas can help to mitigate fires before they get too large to control and we lose forests or create hazardous conditions for people living around the area. This is a clear example of AI assisting in a real-world problem that just isn't feasible for man-power alone to solve.

Get Creative with AI

Practice using AI and unleash creativity with your new collaborator

Copy the prompts below and paste them into ChatGPT, BingAI, Google Bard, or your generative AI tool of choice and let AI work its magic. Share your results with us…and take it a step beyond the initial prompts and see what you can create!

Text-based

• Imagine you woke up with the ability to speak and understand any language in the world, from dolphin clicks to computer code! What would you do with this newfound linguistic superpower? Where would you travel, who would you talk to, and what fascinating conversations would you have?

Visual-based*

• Create a futuristic image of what [enter your favorite city] could look like in the year 3023

Visit us on Instagram and Twitter to see how our examples turned out!

*Visual-based examples are for users of Dall-E, BingAI, or premium users of ChatGPT - may require paid subscription.