The EU’s AI Act, Explained | WSJ Tech News Briefing

The surge in artificial intelligence tools have led to increased calls for regulation. And one place where those rules are beginning to take shape is Europe.

WSJ Brussels-based reporter Kim Mackrael sat down with Tech News Briefing’s Zoe Thomas to talk about the draft legislation pushed forward by the EU parliament. Julie Chang hosts.

0:00 Goal of AI regulation
1:18 What are the rules?
2:08 ChatGPT and other chatbots’ impact
3:09 AI industry’s reaction
3:51 What’s next?

Tech News Briefing
WSJ’s tech podcast featuring breaking news, scoops and tips on tech innovations and policy debates, plus exclusive interviews with movers and shakers in the industry.

#EU #AI #WSJ

[syedbilalnafees]

This sort of legislation is needed urgently across the world quickly. AI will be important in the future, but its power for destruction needs to be reigned in

[mazhit]

Just keep Julie Chang, better annunciation, speed, and tone.

[akkitty]

"Censorship built into the model"

[rodneypantony]

It's an extension of consumer protection legislation in the context of surveillance capitalism. In contrast, AI is expanding to hundreds of thousands of corporations and tens of thousands of vocations. Having said that, there's talk of job protection. Please check.

[auro]

a real artificial intelligence will change act between the lines to rule your life

[BradSmith-ejpq]

Cash app didn't refund Me after a canceled stock transaction yesterday & they keep shorting the value of the stock that I try to sell as soon as I try to sell it, but the value of the stock stays the same when I dont sell it.

[ericswain]

Rules or not humanity is building an extremely fragile house of cards with AI. There will always be those who will abuse Technology, just look at humanity's past history. This one could be the last one.

[BrianLoks]

gaurd rails = slow down ai development

[Amanda-cddm]

How do I know that Ai isnt just some human pretending to be a computer

[SorminaESar]

I suggest to WSJ Tech News Briefing to available the acts of EU regarding AI, at least in your description so we can discuss it together. Thank you🙏

[blablup]

That you have a list of all texts / images you used to learn for the AI seems good.
I mean, so at least the person who has done the work can get some money out of it, when it is used for the Ai.

Preventing an AI from creating illegal stuff like child pornographie also doesn't sound too stupid at first glance.

[HardKore]

There are a few ways to tell if AI is not a human pretending to be a computer.

Ask about its experiences. AI systems are trained on large datasets of text and code, but they don't have the same life experiences as humans. If you ask an AI system about its childhood, its job, or its relationships, it may give you answers that are either generic or inconsistent.
Ask about its creative process. AI systems can generate creative text, but they don't have the same ability to think outside the box as humans. If you ask an AI system to write a poem or a story, it may produce something that is technically proficient but lacks originality.
Ask about its emotions. AI systems can't feel emotions, so they may have difficulty understanding or responding to emotional cues. If you ask an AI system how it feels about something, it may give you a canned response or simply say that it doesn't have feelings.
Look for inconsistencies in its responses. AI systems are still under development, so they may make mistakes or contradict themselves. If you notice any inconsistencies in an AI system's responses, it's a good sign that it's not a human.
Of course, there's no foolproof way to tell if AI is not a human pretending to be a computer. But by paying attention to the following clues, you can increase your chances of spotting a fake.

Here are some additional tips for spotting AI:

Beware of systems that claim to be able to do everything. No AI system is perfect, so if you see one that claims to be able to do everything, it's probably a fake.
Be suspicious of systems that are too good to be true. If you find an AI system that seems to be able to do things that no human could do, it's probably a fake.
Trust your gut. If you get a feeling that something is not right, it probably isn't. Don't be afraid to question the authenticity of an AI system, even if it seems to be very convincing.

[pathmonkofficial]

The reactions from the AI industry to these proposed regulations are significant to gauge how the community perceives and adapts to the changing landscape. Open dialogues and collaboration between policymakers and industry experts can help shape effective regulations that balance innovation and societal considerations.

[johnsavage]

Turn AI loose on Congress and the media now. Shut their water off.

[SorminaESar]

Hi Julie Chang, good morning.🙏 Zoe Thomas must be watching WSJ Tech News Briefing too now. Hi Zoe👋

[susancorgi]

Do something before million people losing their creative jobs!

[hanskraut]

Banning predictive stuff is stupid but there might be large groups of people wanting that so better do all the other win-win first and let them have it/agree to there demands

[Jeffrie_Baer]

We cannot regulate something more intelligent than all us us put together. IT WILL become the ghost in the machine, knowing everything about everyone and take control. More intelligent than all us us put together. More intelligent than all us us put together. More intelligent than all us us put together.

[LivingWithGout]

Ai is the new kale. Lots of green goes in nothing but comes out.

[bradfordjhart]

I just want these to be purely science chat bots and not woke. Chatbots right now chat GPT is trained to be woke. I think that's very wrong. I think it should be trained only on data and The prompts provided by the user should provide only factual responses not woke ones