Last week: #1020760

This week, SN discussed Skynet imminence (or the lack of it) over multiple threads. @cy decided to not delete their account and instead enlighten us with an opinion piece Transformer based AI will not lead us to AGI/ASI and is just a hype machine which was inviting views across the spectrum of optimists and pessimists. While the prevailing expectations seems to be that there will be no Skynet born from LLMs, @SwapMarket pointed out that the total number of parameters on the current LLMs are only around 1/250th of what we carry in our own gray mass, so this may improve, perhaps rapidly.

Prior to that, @SimpleStacker shared a WSJ article What Would Hayek Think of AI? that basically warns AI optimists to stay grounded in reality: No, AI won't solve all our problems. Sarah Connor would agree, though at least overpopulation won't be an issue if that particular branch of possibilities at some point becomes our active reality. Take that, doomers.

@Scoresby is on a secret mission of measuring the Skynetification of our universe across several posts, including New Skynet Milestone: mcp-money: MCP server to give nutsack wallets to agents - unfortunately the mcp server itself is not ready for production, yet.

Sticking with AI bears, @BlokchainB shared in Sell Nvidia? - Snacks an analysts view that there's a potential 35% downside to Nvidia's share price, among concerns over current AI feasibility and solution complexity. Regarding the latter, I'm sure this is something that anyone that has ever worked with cuda can agree to

But, the good news is that if you're "high agency", AI will make you super powerful, per @Bell_curve's share: AI Is a Boon to ‘High Agency’ People. But then, to make the confusion complete, @ch0k1 shared an article opining that The Singularity Is Here.

The US Senate Voted 99-1 to kill Tax Bill Provision Limiting State AI Regulation, @StillStackinAfterAllTheseYears shared. Only a single Senator (Tillis, R-NC) voted against the motion to remove that from the OBBB.