I was actually thinking more like participants who can offer computational power during periods of no use, accrue reputation the more they contribute, and can use that reputation to offload computation to the cloud. If your PC was not powerful enough to run local LLMs, it would need to contribute a lot more, or perhaps it would just be limited to clouds of people running powerful enough GPUs for the local LLMs being used so they could get an added benefit to leaving it running at night.

An LLM model running over across multiple PCs across the Internet would not really be workable, no, but their product is.

I’m still waiting, reddit, why are you still accessible publicly? Second thoughts, spez? Seems this is taking longer than other announcements, like when they locked old.reddit behind a login. I saw the login prompt pop up on desktop once, and never again after that.

Reddit has repeatedly proved that they do not care about their users, regardless of what they say. No one who has left is eager to get back, they are only trying to catch new users which other social networks are much more effective at.

It also allows us to protect access to Reddit’s extremely valuable corpus of data

That’s the real problem, people have to stop giving Reddit their data to monetize before they inevitable kick them to the curve and remove any control they had over their own data.

In another timeline where RAM and GPUs were still largely consumer accessible, I wouldn’t doubt there would already be a SETI-like decentralized alternative for LLMs.

It doesn’t fulfill the same user case. One can be used offline, the other cannot. One can be used without a subscription, the other cannot. One needs to be running constantly to meet targets, the other can be run whenever I want. One has that memory being used by very specific services, the other has that memory being used for whatever I want to use it for. Why should we care about an efficiency factor that really only benefits Sam Altman and ignores all other metrics?

If you think he has a point, I think you are being dazzled by the inherent proposition of an implementation in the ideal egalitarian society, which we do not live in nor is he really interested in providing.

Great, it’s removing itself from a lot of free publicity it was getting from search results and basically conceding them to the AIs spez was always using as a strawman.

Spez has all but basically admitted that reddit is a targeted experience now and that the only posts you will see are ads, echo chambers, and rage baits. It has little to do with its origins now.

Finally, I was getting sick of every nuclear proponent who seemed to discard it.

They are really going to be used for one thing: walking flock surveillance. They are going to be deployed around neighborhoods to see who they can spot that’s latino looking enough or who’s simply on their suspect list. Wanna bet if they will get the local police departments to deploy them for ICE as well? They will spot suspects, build a profile, and when they have enough ICE moves in.

It’s an informed assumption based on historical precedent and the combination of plausible deniability with level of ease to implement that only statistics can really disprove as anything but speculation. There’s also less eyes this time around because independent observers like OSCE have been prevented from participating.

These issues are nothing new, but are so old some of the articles that have discussed them in the past are no longer even accessible:

https://verifiedvoting.org/vote-flipping-and-touch-screen-calibration/

That isn’t really what is at risk, that’s just a MAGA poll worker deliberately sabotaging the touch screen calibration. It might be an attempt at election theft, but what Trump is seeking to do is much more covert and you won’t even notice it. They will try to see if and where they can flip the tallies and mess with the paper trail if there is any, and that’s only on one of the levels he will be trying to steal the midterms.

Seems a bit like anti Local LLM propaganda to begin getting the ball rolling on subscription based LLMs being the only LLMs people have access to.

They’ve been doing everything to try to keep the technology from the general public, from helping to condemn any use of it (even that not owned by corporates and dedicated to allowing open access to it, which many people willfully play along to) to passing it off as a threat that needs to be under control (not by governments, just their own PR stunts with certain mouthpieces). This with the price of computers skyrocketing is helping keep away any possibility of this technology being owned by anyone but the major circlejerking multinational pricefixing corporations.

Or maybe it’s just a study.

This probably has a wide latitude of success depending on whether it’s a high trust or low trust society.

Then I must really just fall left of the bell curve. It is a flawed intelligence, and intelligence does not mean wisdom, nor does the power of knowledge have anything to do with the power of feeling or power of will. We are prediction machines as well. The nature of our ego makes us prone to be prejudiced, specially with metaphysical manifestations we can’t quite grasp. Sometimes there is no greater danger than underestimating your opponent while overestimating your own nature.

Oh, to be spurned by Dario…

The alternative is for the AI to probably even know more about you so it knows more about the context. I tried to reproduce it on Gemini, same thing - except I clarified the context a bit by saying ‘I was just searching online for the reference “hes never coming over dario” to find the article that used it’ and it linked me to a repost of this very submission. A Google search did not automatically invoke the AI overview.

Honestly, some people would probably welcome this, but if a big corporation puts out something that acts like a therapist while harvesting people’s data, it should be liable for its mistakes and the data gathering as well.

There’s more than one movie about aliens who use global warming to wipe out humanity. AI data centers seem to be doing just that in more than one way. No need for an alien invader though, even a primitive species adapted for all the nooks and nuances of the environment is far more adept at it, it just needs to be diseased.

The real danger isn’t a smart AI - the smartest people don’t end up being in charge of our societies. The real danger is having enough control to be able to con most of our society, and seeing how much our societies are willing to continue participating in clearly compromised special interest social networks, I don’t have a lot of faith. Barely a difference if humanity is replaced or just led to the lowest common denominator of exploitation at that point, and that’s where we are heading politically.

Gambling implies there’s still the concept of luck and not of insider trading.

They’ve already disabled old.reddit for me. That makes it unusable, and thank you, Reddit. I actually used a domain blocker to block reddit, but generally would still be tempted to peek. Now, that is no longer the case. I, for one, am wholly in support of Reddit’s new anti-advertising stance!

They are in the business of attracting new users into their new algorithmic engagement hellhole now. Show any propensity of interest, and you will get sidetracked to the most godawful side-communities that seem to have emerged to engage as many victims as possible. They do not want to focus on their old users as anything less than the content they already made that makes reddit show up as free advertisement to their new base in search engines. It’s all a game of “it’s the algorithm’s fault so you can’t blame us” now.

It certainly is using generative AI methods given Jensen is calling it that. People have their own issues with things, but IP theft is the biggest for me given it shows the disturbing amount of leeway people developing have been given to break the law.

I wouldn’t have called this generative AI, but Jensen did. Great for stills, uncanny valley for motions. People are claiming it generates completely new images, but in this instance it keeps the same geometry and texture and just processes motion and color vectors to create a hyper-rendered version of the characters.

Ignoring the obvious problems of the hardware it requires and supporting the AI bubble feeding monopoly that is NVIDIA, it is interesting technology that doesn’t actually seem to act as a medium for IP theft, my beef with what I call AI slop. It might only be practically good for photo mode in games, but it will be interesting to see how it works out. It could kickstart interest in making Let’s Play in a more Machinima style.