AI regulation is coming governments Around the globe have been actively Discussing how they can control this new Technology and politicians in the eu's Parliament recently approved the world's First AI rules if history is any Indication the eu's AI rules could Become the standard for every country And region reviewing these regulations Is therefore required to understand What's coming so today I'm going to give You a brief history of the eu's AI rules Tell you what they say explain what they Could mean for you and assess whether AI Regulation is even possible Used AI rules have their roots in the Block's digital decade initiative the Second phase of which began in 2020 Around the time the pandemic got Underway Like all the other initiatives that Nobody voted for the eu's digital decade Initiative seeks to completely transform The continent by the year 2030. if You've watched our video about the Online censorship laws being introduced Around the world you'll know this 2030 Date is everywhere because of the un's Sustainable development goals or sdgs Which every country is supposed to meet By 2030. let's just say that many EU Politicians have drunk that Kool-Aid That said not all the eu's digital Policy proposals have been dystopian one
Example is the digital markets act which Forces big tech companies to allow you To change the software on their devices As a result of this rule you will soon Be able to say uninstall iOS and install Android on an Apple device The eu's AI act seems to be another Example of surprisingly reasonable Regulation the only caveat is the AI Act Is a regulation meaning that it will Override any AI regulations that European countries already have in place Without the input or approval of their Citizens that's the EU for you anyways The text of the AI Act was published by The European Commission in April 2021 Long before open AI had introduced chat GPT For context only the European commission Can table policy proposals in the EU These proposals are then debated by the European Parliament until they are Passed this is basically what's been Happening over the last two years but There have only been a few amendments so Far most of these have had to do with AI Technologies that have emerged or become Popular since then such as generative General purpose AI which of course Includes chat GPT on that note openai Has reportedly been lobbying EU Politicians to ensure that their AI Rules don't overburden the company some Of you may remember seeing a few
Headlines about CEOs Sam Altman Promising to pull the company out of the EU if these proposed AI rules don't Change according to Time Magazine open AI lobbied EU politicians to drop the High risk label for quote Foundation Model generative AI Tech Technologies Such as chat GPT and Dolly It seems Sam succeeded the final text Approved on the 14th of June ensures That open ai's Technologies operate with Less scrutiny speaking of which some of You might know that usdc stablecoin Issuer Circle has also been influencing The EU when it comes to crypto Regulation if you watched our most Recent video about Mica you'll know that It contained significant restrictions For stable coins that aren't set up like Usdc funny that Now to clarify the AI act still has a Few more Hoops to jump through before it Becomes full-blown regulations if we Don't reach some sort of AI Singularity That requires a last minute change the AI act will be finished sometime early Next year even then it may not actually Come into Force until 2026. in any case The contents of the AI act are where Things get interesting for starters It'll apply to any individual or Institution that's developing or Offering AI Technologies used in the EU Even if the individual or institution is
Not located in the EU this explains why Openai has been lobbying so hard The ACT puts all AI Technologies on a Risk Spectrum with lowest risk Technologies being mostly unregulated And the highest risk Technologies being Entirely prohibited if you're wondering What high risk means well the original Text provides a clear definition the EU Could sure teach the SEC a thing or two Jokes Aside high risk AI Technologies Are quote those that have a significant Harmful impact on the health safety and Fundamental rights of persons in the Union It appears that the updated text Includes harm to quote the environment Democracy and the rule of law now all a Bit vague if you ask me not surprisingly The list of prohibited and high-risk AI Technologies has expanded because of This expanded definition the prohibited List includes AI use in biometric Systems in publicly accessible spaces AI Use in predictive policing AI use in Emotion recognition systems and AI use In Social scoring the high risk list Includes AI use in critical Infrastructure such as Transportation AI Use in education especially scoring AI Use in essential services including Issuing loans AI use in employment and Hiring and AI use in Justice and Democratic processes note that all these
AI uses will be permitted obviously Anyone offering these high-risk AI Technologies will have to comply with a Long list of rules including providing a Detailed risk assessment ensuring that Their data sets contain no bias keeping Track of everything they do and Providing clear information to anyone Using their Technologies now this ties Into the requirements for low-risk AI Technologies which are limited to Transparency and labeling obligations Users of these 10 Technologies just need To know that they're interacting with an AI the foundation model generative AIS Fall into this category albeit with a Few more requirements The EU also outlines a third category of AI Technologies which offer no Measurable risk these include AI use in Video games and spam filters logically These kinds of AI technologies have no Requirements now the eu's website notes That most of the AI currently used in The EU falls into this third category Now as you might expect there's a long List of fines for those who fail to Comply with the requirements of the AI Act the most severe sanctions apply to Those who operate prohibited AI Technologies the fines can be as high as 40 million Euros or seven percent of Annual revenue whichever is higher a Failure to comply with data related
Requirements such as not using Copyrighted content or transparency Related requirements such as not Labeling AI content could result in Fines of up to 20 million Euros or four Percent of annual revenue whichever is Higher a failure to comply with any of The other requirements in the AI act Could result in fines of up to 10 Million Euros or 2 percent of annual Revenue whichever is higher as a bonus Providing incomplete or incorrect Information to Regulators can result in Fines of up to 5 million Euros or one Percent of annual revenue whichever is Higher say this sounds potentially Pretty lucrative for the EU In all seriousness you should know that The AI act scrutinizes any AI use cases Involving Biometrics or facial Recognition it seems that EU politicians Want to ensure that citizens in the Block won't be governed by AI in any way This is certainly good news but the AI Act could still be bad news in practice Recall if you will the definition of a High risk AI technology quote those that Have a significant harmful impact on the Health safety and fundamental rights of Persons in the union including quote the Environment democracy and the rule of Law again this definition is excessively Broad and arguably problematic for Starters the definition of significant
Harmful impact is open to interpretation And it's guaranteed that this definition Will be stretched when it comes to Things like the environment democracy And the rule of law its area really Similar to the wording in the Digital Services act which legalizes online Censorship case in point there have Recently been a series of Articles Asking whether AI is harmful to the Environment due to its energy Consumption and such if EU bureaucrats Ever started butting heads with foreign AI firms they could simply label them as Environmentally harmful and kick them Out the same goes for the Democracy Aspect the most recently approved text Of the AI act includes Provisions that Effectively apply a high risk label to Any social media algorithms associated With elections any social media platform With more than 45 million users will Fall under scrutiny consider a scenario Where an anti-eu candidate is running For president in one of the eu's Countries EU bureaucrats could prohibit Social media platforms from allowing This candidate's content to go viral by Labeling it as being harmful to Democracy some would say this has been Happening in the EU already The rule of law part could have a Similar outcome this would presumably Include any AI technology that is in any
Way involved in creating skepticism Around the EU and its often undemocratic Regulations it's important to remember That the top priority of these Institutions is to protect themselves And it's not just the institutions Either it's hard to see how significant Harmful impact could be measured at an Individual level what counts as Significantly harmful varies from person To person setting personal harm as the Threshold means lowering it to the level Of the most sensitive person in society Now to be fair it's more than likely That more than a few people would have To be harmed for the EU to determine an AI to be significantly harmful Some would say that we live in a very Sensitive society and that's putting it Lightly lots of people also exaggerate Or even fake their personal harm for Attention The Practical effect of this is rules And regulations that only make sense to A small minority of people usually the Minority that you see on the mainstream Media for everyone else it results in Unnecessary restrictions and in the case Of AI in the EU that means falling Behind in a race it can't afford to lose To their credit EU politicians have Proposed creating a regulatory sandbox For new AI technologies that allows them To assess their risks more objectively
The problem there is that this means the EU can control which AI Technologies are Allowed never mind that its assessment May not be objective And don't even get me started about how Much control the EU could have over high Risk AI Technologies such as those Involved in education something tells me That there are more than a few EU Bureaucrats who would love to leverage These tools to make sure that every Child grows up loving the Union So this brings me to the big question And that's if it's even possible to Regulate AI well the short answer is yes But it requires either controlling the Software or the hardware in other words It requires controlling the internet Where this software is used or the Advanced microchips that make the Software possible controlling the Internet is easier said than done but it Seems to be on the roadmap of almost Every government what they desire is for Everyone to require a digital ID to Access the internet so that everything Is closely tracked and controlled that Would make it impossible for rogue AI to Arise it's also possible that Governments would try and Implement an Alternative internet which is controlled At every layer believe it or not but the Internet as designed today is fairly Decentralized however one of the points
Of centralization is internet service Providers and this is where governments Could take control if you watched our Two-part video series about the world Economic forum's Cyber attack Simulations you might recall it's Speculating that the end game of a Global Cyber attack could be to force Everyone to switch to a single internet Service provider such as starlink this Would likewise make it impossible for Rogue AI to arise when it comes to Hardware meanwhile it's easy to forget That AI doesn't work without it and even Easier to forget that most microchips For advanced AIS are currently made in Taiwan given the geopolitical tensions In that region the cessation or even Just the disruption of production could Stop AI dead in its tracks now it may Sound crazy but this is exactly what the US and the EU are planning for the fact That both are rushing to create their Own Advanced microchip factories is a Tacit admission that some kind of Conflict in the Taiwan Strait is not a Question of if but when the answer could Very well be sooner than we think Speculation aside the fact of the matter Is that the US EU and others will soon Have their own microchip manufacturing Facilities we could easily see them Implement Advanced security controls That would only allow AI related chips
To function if their purchaser has Received regulatory approval some would Say that this is also a question of when Not if there's been an undeniable Trend Towards extreme centralization in recent Decades and the result will be the Centralization of as much software and Hardware as governments can trick or Force their citizens into accepting now With all that said the EU has taken a Third approach and that's to focus on The data by forcing anyone working with AI Technologies to provide their data to Regulators it becomes impossible for AI To evade regulation now this is clever But comes with lots of loopholes that Other EU regulations are trying to close Put simply the effectiveness of the eu's AI regulations ultimately depends on Whether the union can maintain control Of the data that AI uses to operate now It's too soon to say if this Innovative Approach will work but if it does then You can bet that we'll see other Regulators elsewhere take the same Approach then again it's possible that Some or even most countries will try and Postpone AI regulations as much as Possible this is because most Governments want AI to evolve as quickly As it can the faster it evolves the Bigger the advantages they have this is A big part of why Regulators have been Slow to rein it in the only reason
They're starting to step in now is Because they're realizing that allowing AI to evolve unchecked could backfire Spectacularly Besides the possibility that AI could Somehow become sentient and start Breaking stuff many countries are likely Using AI against their geopolitical Opponents this use is probably in the Form of information Warfare primarily Via social media and it's not just Countries either as domestic tensions Increase the use of AI for propaganda Purposes will become more common some of You will have heard about the AI Generated ads recently used by a Politician in the United States it is Going to be fascinating to see how Governments balance allowing AI to Evolve while preventing it from evolving Into something they don't want I suspect That it's going to look like what we're Seeing with crypto regulations today but The stakes will be much higher it's not Just money it's information as well make No mistake you should be keeping track Of AI developments and regulations it Might not seem like it now but both are Going to have a profound impact on every Aspect of Our Lives just like the Internet and crypto it's going to take Time for that to happen but you best Believe that it's coming We'll be sure to keep you updated on
Both topics in addition to your Regularly scheduled crypto and macro Content and if you happen to be worried About whether AI will take your job be Sure to check out our video about that Using the link in the description And that is all for today's video if you Found it informative help others find it By Smashing that like button if you want To make sure you don't miss the next one Subscribe to the channel and ping that Notification Bell better yet take a Second to share this video so that more People are in the know And if you happen to be looking for a Hardware wallet to keep your crypto safe Or an exchange that won't shaft you with Fees head on over to the coin Bureau Deals page it has got the biggest Discounts on the best crypto wallets and Up to forty thousand dollars of bonuses On offer at the best crypto exchanges The link will be in the description Thank you all for watching and I'll see You next time Thank you [Music]
Coinbase is a popular cryptocurrency exchange. It makes it easy to buy, sell, and exchange cryptocurrencies like Bitcoin. Coinbase also has a brokerage service that makes it easy to buy Bitcoin as easily as buying stocks through an online broker. However, Coinbase can be expensive due to the fees it charges and its poor customer service.