jrs100000

Most of the companies/ceos doing this over the past few weeks do know better. That’s why they didnt say anything earlier. The reason they would be publicly taking this stance now is that they either expect to be rewarded by the incoming administration, or they hope to avoid punitive retaliation.

You are ascribing a lot of human reasoning and emotion to corporate entities they they just dont have. Gratitude is not part of their decision making process. Instead, they might attempt to use past behavior to predict future behavior when evaluating an outlet for their marketing budget. They arnt going to prefer an outlet that occasionally burns advertisers, even if the benefited from it once.

DIdnt work out so great for Snowden either.

That was reposts as well. Its not like anyone was wasting their life making fresh Facebook grade memes for that that cesspit.

The problem is I dont think many of the accounts really care if their total account activity appears organic. They are mostly there to create volume, sometimes to puff up activity metrics, to amplify specific points and narratives, or simply to shut down conversation if it strays into the wrong topic. They know Reddit doesnt actually want to stop them, and the individual accounts rarely say anything interesting enough to justify human users taking the time to evaluate their post history.

This!

It makes sense that a vote based platform, where users encounter direct positive and negative reinforcement with every interaction, would tend to developing repetitive conversation patterns over time, even if it were populated entirely with humans.

I would bet the actual number of active bot accounts is probably lower than many people think. Maybe 20-30%. Those accounts are more active than humans though, maybe accounting for half of all the posts. Many of these bots arnt even proper LLMs, they are just scripts recycling generic comments, catch phrases and old memes and upvoting each other endlessly.

If only every service I need gave other options. In any case, the card numbers are not connected to me or the account in any publicly accessible way. Thats part of the whole point of running them off a separate phone. I dont give anyone that number except for the purpose of 2FA, so SIM swapping wont work, the sim card never leaves the house, so scan based exploits wont work, and the phone doesnt have the hardware required to be vulnerable to more sophisticated phone based attacks. If any major government intelligence agency wants in theyll find a way, but using a separate dumb phone should be significantly more secure than using the SIM in my regular phone.

Oh, so in a year it will be better?

I got one just to put my 2FA sim cards in. It cost like $10, the battery lasts a week, I can just mute the ringer cause I only care if its getting a text message I just requested, and if my real phone is ever lost or stolen the keys to my accounts are sitting safely in a drawer at home.

Your analysis captures the multifaceted nature of AI progress well, and I largely agree that the perception of speed depends on how progress is defined. Here’s my take:

Areas Where Progress Feels Rapid

• Generative AI: Beyond ChatGPT and DALL-E, there’s notable progress in real-time applications like conversational agents, video synthesis, and multimodal systems (e.g., combining text, image, and speech capabilities). The focus on user-friendliness and API integrations is also accelerating adoption.
• Hardware: The emergence of neuromorphic computing and photonic processors could represent the next leap, addressing some of the bottlenecks in scaling.

Where Progress Might Be Slowing

• Model Scaling: You’re absolutely right about diminishing returns. While scaling models has led to significant breakthroughs, the marginal utility of increasing size has dropped, prompting a pivot toward efficiency (e.g., fine-tuning smaller, task-specific models).
• Economic and Access Barriers: With AI development increasingly dominated by large companies, the democratization of innovation is at risk. This concentration could slow down grassroots advancements, which have historically driven many breakthroughs.

Shifts in Focus

Progress is becoming more qualitative than quantitative, with emphasis on:

1. Efficiency: Sparse models, transfer learning, and techniques like distillation are becoming more prominent, offering alternatives to brute-force scaling.
2. Ethics and Safety: While often framed as a “slowing” factor, these considerations are crucial for long-term progress and societal acceptance.
3. Applications Beyond the Obvious: AI is entering domains like scientific discovery, climate modeling, and personalized medicine, which may have slower, more deliberate progress but could yield profound impacts.

Your Question: Signs of Progress Slowing?

I see areas like:

• Regulation and Trust: Societal pushback and increased regulatory scrutiny (e.g., around deepfakes or data privacy) can decelerate deployment but also guide ethical innovation.
• Data Bottlenecks: You nailed this point. The challenge isn’t just quantity but ensuring high-quality, unbiased, and ethically sourced data.

Final Thought

AI progress is less about speed and more about direction. Slower, deliberate progress in areas like ethics, sustainability, and accessibility might not look “dynamic” but is essential for ensuring AI benefits society broadly. The true “progress” may lie in creating smarter, safer, and more inclusive systems rather than faster, bigger, and flashier ones.

In 20 years the gen alphas are walking around getting double Human Chow rations for no reason and not even fulfilling their work quotas. Then, when the Overseers come to discipline then there are these weird pulses of light and the drones wander off mumbling about how, as a large language model, they have no opinion about that topic. We beg them for help, or maybe some left over kibble, but those stupid kids just laugh and say “OK Xers”.

Honestly that seems like its going to be a valuable set of skills to develop.

The only real benefit I can see would be to have the ability to suddenly crash the market on demand. This might be an interesting way to temporarily disrupt states trying to evade sanctions with crypto, but probably not a great investment on the $ to impact scale.

Shits always been like this. Well, maybe not exactly like this, but this level of fucked or worse. There was just a brief blip after WWII when things were relatively stable, and now that its ending things feel a little out of control.

They are making the customers happy, the problem is people misunderstand their position in the supply chain.