YouTube Deep Summary

[00:00] Hey everyone, welcome back to the [00:01] channel. My name is John and this is your modern tech breakdown. Today I'm covering another attempt by the California legislature to regulate the emerging AI industry. Let's jump into it. [Music] All right, to start off, let's just recap from last year in case you missed it. Last year, SB 1047 was passed by the [00:30] California legislature, but was [00:31] ultimately vetoed by Governor Gavin [00:33] Newsome. That bill was an attempt to regulate AI companies in the state of California. Now, the same legislature is back, this time with Senate Bill 53 that would require AI companies to publish safety and security reports for chat bots because apparently these things are super scary. But let's get into the bill details a little bit. As it stands now, and I should point out this bill has been amended multiple times already, so it's likely to change again. But as of [00:59] now, the bill would create a new [01:01] reporting requirement for AI companies, [01:03] including documentation of what data was [01:06] used to train the model. As an aside, this alone might kill the bill. AI companies definitely don't want to give away any shred of information about how they train their models. It also requires disclosure of safety and security protocols and testing procedures. Basically, it's the start of the construction of an AI bureaucratic machine. It starts with reports, but we [01:26] all know it's not going to end there. That much is obvious. It would also create a new group within the California Government Operations Agency that would be responsible to create an AI called Cal Compute that is safe, ethical, equitable, and sustainable. I think they hit all the buzzwords right there. What do any of these things actually mean with regard to the development of AI? [01:47] Your guess is as good as mine. These terms are so vague that they can be made to mean basically anything that bureaucrats in Sacramento want them to mean. Honestly, it sounds like a politician's kickback scheme to me. Spread out some of the public's money over some groups that are favored by the people in charge. And lastly, the bill also protects whistleblowers of AI companies when they think their employers products pose a quote critical risk. And we've already seen employees [02:11] leave some of these large AI companies [02:13] and make silly statements about how [02:14] they're worried about AI. And which again, I will mention that these AI products are more or less just chat bots at this point. They're not going to hurt anyone yet. I think we're putting the card before the horse just a little bit. But all this talk about safety in this bill got me thinking, where have I heard about AI safety before? Isn't there a [02:32] company out there that talks non-stop [02:34] about AI safety? Ah, yes. It's anthropic. So, that got me digging a little bit. Who is donating to the sponsor of this bill, Scott Weiner? [02:43] Well, I noticed a company in the list [02:45] here called SV Angel LLC. So, who is SV Angel? It's an angel investor vehicle for Silicon Valley companies founded by Ron Conway, who just so happened to have been an early investor in Google. And his company SV Angel seems to be pretty active campaign donator, but conspicuously only on one side of the aisle. And who else has SV Angel invested in? Oh my, lookucky there. It's [03:11] Anthropic, the company that always seems [03:13] to be popping up in all these silly [03:14] discussions about AI safety. Wow, what are the odds that an angel investor in Anthropic also donated to the campaign of the California senator that introduced this AI safety bill? I mean, what a coincidence. But in all seriousness, this is just gross to me. Anthropic investors are manipulating the California government to enact AI regulation on their behalf. They're [03:35] trying to build barriers to entry so [03:36] that they can have a monopoly on this [03:38] emerging industry. It's just really gross. Uh but I'm not surprised. This is par for the course. But what do you think about it? Leave a comment down [03:45] below. As always, thanks for watching. Please like and subscribe and I will catch you next