AI bosses call for industry slowdown as public concerns grow | BBC News

About a year ago, when AI started becoming significantly more capable, I looked into what Goldman Sachs thought about the potential for job losses. At the time, I saw estimates suggesting that around 400 million jobs could be affected over the next 10 years. More recently, when I asked the same question, the figure I was given was closer to 80 million.<br><br>I follow different content creators who focus on AI, automation, surveillance, and the potential impact of these technologies on society. If, in the United States alone, around 1.6 million jobs are already being lost or significantly affected, there will clearly need to be retraining and re-employment on a large scale.<br><br>But this raises a bigger question: <b>*At what level of job displacement does the economy become seriously destabilised? Is there a point where the loss of employment and consumer spending could cause an economic collapse?*</b><br><br>I am typing this myself, and I will also use AI to help develop and communicate these ideas. That is part of the contradiction we are facing: AI can make individuals more productive while simultaneously reducing the need for human labour.<br><br>I am also concerned about the development of increasingly autonomous AI systems, including systems connected to drones and other physical technologies. I believe development needs to be accompanied by much stronger testing and safety controls.<br><br>Rather than allowing increasingly powerful AI systems to be tested primarily in the real world, there should be dedicated, controlled environments where systems can be evaluated safely. We need something comparable in principle to a highly controlled laboratory environment: data can be introduced, systems can be tested under controlled conditions, failures can be studied, and safeguards can be developed before deployment.<br><br>I believe the major AI developers should seriously consider creating an international AI testing facility—or network of facilities—where advanced AI systems can undergo standardised safety testing before being deployed at scale.<br><br>Ideally, this would involve governments and organisations from around the world. AI development is no longer something that affects only one country or one company. The potential economic, social, security, and surveillance implications are global.<br><br>The goal should not necessarily be to stop technological progress permanently, but to make sure that the development of increasingly powerful AI happens alongside rigorous testing, international cooperation, and meaningful safeguards.<br><br>a super lab is needed it doesn&#39;t matter who&#39;s cuda it runs on .before it is to late

Just wanna note here, if a tech bro tells you something BE SKEPTICAL OF THE MOTIVATIONS.<br><br>Remember all the promises they&#39;ve made over the years, like how were gonna colonize Mars by now? The failure of the hyperloop? How bitcoin was going to replace real money? They lie and exaggerate, they&#39;re salesmen and that&#39;s what they do..<br><br>This is likely yet another scam by tech bros, they realized they went too fast for a product that isn&#39;t making money and is taking a lot of investments that they&#39;ll need to pay up soon and the investors they do have are slowly waking up to the fact that this crap isn&#39;t making money. Just be more skeptical.

Watched an interesting interview on MSNOW yesterday where someone was saying that the true motive for ringing the alarm bells by the big players is that they are trying to get politicians to slow down the open source projects. This person was saying that the big players have invested so much, they wouldn’t see nearly as much in returns if they had to compete with open source projects. Some of these big companies are just about to be listed publicly and need to show their potential investors that there is some money to be made but 99% of the tasks you would typically give to AI can be done by open source agents. 🤔

This is nothing but a stunt to ramp up Anthropic&#39;s IPO in October and it is complete hogwash. We have heard it all before. The real danger from the leading AI companies is that they are all billions of dollars in debt, and are looking for a public bail out. They are doing it in a really subversive way by listing on the S&amp;P 500, which recently changed the rules around Index Investment funds which previously did not have to buy stocks in listed companies for at least a year, I believe. It is now just 18 days. So what that means is that any pension fund invested in the S&amp;P 500 will have to buy stocks in these companies - they are compelled to. And that is the real danger - money and the impact that just one of these companies going bust could have on economies, including pension funds.