AI & Robotics News and Discussions

[firestar464]

[firestar464]

[weatheriscool]

[firestar464]

[spryfusion]

[firestar464]

https://www.wheresyoured.at/the-haters-gui/

I may not be a contrarian, but I am a hater. I hate the waste, the loss, the destruction, the theft, the damage to our planet and the sheer excitement that some executives and writers have that workers may be replaced by AI — and the bald-faced fucking lie that it’s happening, and that generative AI is capable of doing so.

While there are a few sadists in the Trump administration who would enjoy the loss of jobs without any form of restitution, the vast majority of singularity supporters are in reality aiming for a state of post-scarcity, where no one HAS to work to live. It is disingenuous to frame those calling for the abolition of work as sadists.

I’m tired of reading people talk about how we’re “in the era of agents” that don’t fucking work and will never fucking work.

where my goalpost moving buddies at? Ever heard of the objective improvements in AI capabilities? No, because you have already resigned yourself to the conclusion that AI will never succeed at anything and will deny improvements when faced with evidence of them. Sure, current agents are shit, but are you seriously expecting things to stay that way? Note that later in the article, the author criticizes current "agents" as not being true agents; he should make up his mind, as you can't criticize future real agents based on the pitfalls of fake agents. Now if you want to argue that true agents are impossible, then argue that instead.

The central point of this essay is that no one is profiting from AI and will never profit.
Companies are hoping for future profit. Now now, sure, that can be a sign of a bubble, when investors eat losses in hopes of future return on investments of dubious quality. However, performance on benchmarks designed to measure reasoning ability continues to improve. Even subject-matter experts clearly critical of the world's AI companies (that is, Kokotajlo and his friends, as well as Geoffrey Hinton, a pioneer of the field) go so far as to warn of human extinction from misaligned ASI. But why listen to them when you can just dismiss an entire field of science because you don't like it? Sure, we'll just live in a Y2K trap! Forever! There will definitely be no significant technological breakthroughs this century!

The discussion of the term "AGI" is an attempt to suggest that Large Language Models can create conscious intelligence, a fictional concept that Meta's chief AI scientist says won't come from scaling up LLMs.

This isn't correct. No one is talking about consciousness. Sounds like cherrypicking to me when suddenly the author cares about experts. And of course he had to indirectly reference neurosymbolic AI by citing LeCun. It is for this reason that I've been Pavloved into cringing when someone says "neurosymbolic AI" or even hints at it. The concept itself is fine, but it's often mentioned by one of those "AI is really bad and will ruin everything but at the same time it's somehow also useless" people that are saying it. LeCun has become the go-to guy for those in denial, and it's really annoying.

The use of stories about models "lying, cheating and stealing to reach goals" or "stop themselves being turned off" are intentionally deceptive, as these models can (and clearly are) being prompted to take these actions.

As far as I know, this is a blatant lie.

If you're going to say I'm wrong, sit and think carefully about why. Is it because you don't want me to be right? Is it because you think "these companies will work it out"?

No. I'd take the world you're predicting over either outcome of AI 2027. You're still wrong regardless.

The real bubble will happen when one lab achieves AGI/ASI and renders all the others as well as various AI startups useless. Companies will die, tech won't. After all, we're still using the internet. He is right that there will be a bubble. He is wrong to conflate that bubble with the tech.

I will admit that the whole AI boom was way too early. ChatGPT 3.5 was mostly a parlor trick without any tools to make it useful.

In my opinion, the most irritating thing about this article is not the points I disagree with, but rather its tone that frankly lacks humility. Gives such "smug redditor" vibes.

I am disgusted by how many people are unwilling or unable to engage with the truth, favouring instead a scornful, contemptuous tone toward anybody who doesn't believe that generative AI is the future. If you are a writer that writes about AI smarmily insulting people who "don't understand AI," you are a shitty fucking writer, because either AI isn't that good or you're not good at explaining why it's good. Perhaps it's both.

Maybe some self-awareness here would help? The irony is striking. And yes, I am capable of criticizing both you and weird hypesters on r/singularity.

[firestar464]

[weatheriscool]

. Humanoid and Bipedal Robots (Advancing Rapidly, Some Available):

This is one of the most exciting and headline-grabbing areas. While many are still prototypes or in limited deployment, some are becoming available for purchase by businesses or specific customers.

Tesla Optimus (Gen 2/3): While commercial sales are projected for 2026, Tesla is rapidly scaling up production for internal use and demonstrations. Optimus Gen 3 is expected to feature upgraded hands, improved water resistance, and more natural gait. Tesla aims to build thousands of units in 2025.


Agility Robotics - Digit: Designed for logistics and warehouse work, Digit is a bipedal robot that can handle tasks like moving totes. It's in commercial pilots with companies like Amazon and BMW.

Figure AI - Figure 02: Another prominent humanoid designed for industrial tasks, with BMW as a key partner for deployment in manufacturing facilities.

1X Technologies - NEO: This humanoid is gaining autonomy with its new Redwood AI model, enabling it to perform household tasks like laundry and navigation within home environments. It's aiming for general-purpose applications.

NEURA Robotics - 4NE1 (3rd Gen) and MiPA: NEURA Robotics is launching its third generation 4NE1 humanoid and the MiPA cognitive household and service robot.

Apptronik - Apollo: Designed for industrial settings, Apollo combines robust hardware with sophisticated AI for intricate tasks and heavy lifting.

Unitree Robotics - G1: A more affordable and agile humanoid, optimized for rapid movement and compact environments. It's available for purchase at a relatively lower price point (~$16,000 USD).

Hexagon - AEON: A newly launched humanoid robot specifically designed for industrial applications to address labor shortages.

Wandercraft - Calvin: Unveiled as a new industrial humanoid with a partnership with Renault, hinting at future factory integration.

KEENON Robotics - XMAN-F1: Debuted in July 2025 as a bipedal humanoid service robot, showcasing role-specific embodied AI solutions for scenarios like medical stations and lounge bars.

AgiBot - A2: Touted as one of the most advanced humanoid robots you can buy right now, with mass production starting in October 2024. It features high AI versatility for tasks like threading a needle and is being used by companies like BYD.

Engineered Arts - Ameca: Primarily for social interaction and customer service, known for its lifelike facial expressions and advanced conversational AI. Available for specific applications.

Sanctuary AI - Phoenix: Focuses on creating seamless human-robot interaction for complex tasks, suitable for service and collaborative work.

[spryfusion]

Project page: https://3d-models.hunyuan.tencent.com/world/
Demo: https://3d.hunyuan.tencent.com/sceneTo3D

[wjfox]

[zzz]

https://3d-models.hunyuan.tencent.com/world/

[zzz]

Those that remember...

[firestar464]

[spryfusion]

[firestar464]

lmfao

[firestar464]

no big releases tomorrow

[firestar464]

[weatheriscool]

[weatheriscool]

[weatheriscool]