It’s a good time to be a AI user

[Mr Positive]

DeepSeek arrival has once again reset AI game and now expect launches even faster more improvement and deep deep reasoning models crazy times!!!

 

[bloopson25]

are you running something with it?

 

[ZinyProxy]

DeepSeek arrival has once again reset AI game and now expect launches even faster more improvement and deep deep reasoning models crazy times!!!

How are you testing it to compare to Chatgpt ?

 

[tidderautomate]

I checked it, and it says 'service unavailable.' Is there a third-party API provider?

 

[peanut1994]

‌‌It directly reduced the cost by 10 times!! Astonishing! I am currently satisfied with r1!

 

[peanut1994]

I checked it, and it says 'service unavailable.' Is there a third-party API provider?

<https://fireworks.ai/ Here you can use Deepseek r1 via API, and it's very fast!

 

[Builder Dave]

I can't wait for proper AGI to ruin my livelihood It's going to be fascinating!

 

[koxokot890]

I can't wait for proper AGI to ruin my livelihood It's going to be fascinating!

What will be fascinating is that the world does not turn into a kind of neo-technocratic feudalism.

It is worth noting that it only takes a few individuals to create a strong disruption in the system, so there is no need for everyone to have an AGI in their pocket.

The biggest impact of DeepSeek R1 is its low cost; how many years do you think you are from the inevitable?

 

[AleG]

I've been coding using Claude, Deepseek and now ChatGPT o3 mini.
I never thought I would be so good just by knowing the definitions, attributes, variables, etc.

 

[india]

AI war will benefit many and will also get many people out of job.. So its good to be AI user every day. Deepseek is very impressive.

 

[prime8]

I am starting to play with all the AI tools available and trying to automate most of the things. The future would be a world of AI, and only a fool wouldn't learn and master them.

 

[Builder Dave]

how many years do you think you are from the inevitable?

Proper AGI is very far away in terms of technological progress left to make. Even experts are not sure whether AGI is even feasible using the current AI architectures. We probably need a few technological leaps to get there and this is not something that's easy to predict. It could all happen almost overnight, or it could take a life-time.

Some futurists of the past would be surprised that we don't have quantum computing or nuclear fusion yet. It was actually a surprise for many that text and image generation came before basic logic or reliable self-driving cars. All the AI progress we have seen so far is based on the transformer architecture. Maybe there are other architectures to come. What I'm pretty confident of is that the next seismic shift forward will also come as a surprise and it will not be exactly as expected.

 

[Don Kalbre]

New DeepSek model are crazy. There are also some new other LLM models which even beat DeepDeek in some benchmarks tests.

 

[koxokot890]

Proper AGI is very far away in terms of technological progress left to make. Even experts are not sure whether AGI is even feasible using the current AI architectures. We probably need a few technological leaps to get there and this is not something that's easy to predict. It could all happen almost overnight, or it could take a life-time.

Some futurists of the past would be surprised that we don't have quantum computing or nuclear fusion yet. It was actually a surprise for many that text and image generation came before basic logic or reliable self-driving cars. All the AI progress we have seen so far is based on the transformer architecture. Maybe there are other architectures to come. What I'm pretty confident of is that the next seismic shift forward will also come as a surprise and it will not be exactly as expected.

Interesting point of view, although I believe that perhaps we may not achieve AGI before a mass extinction. If we define AGI as a super-intelligent entity capable of accomplishing in seconds what would take an entire civilization millennia, perhaps a natural barrier of the universe to prevent such anomalies would be the self-annihilation of any species before reaching such development.

This could partly explain why immortal machines are not out there conquering planets and why there is no sign of life from other civilizations out there. I mean, if the universe operates in cycles, including the life cycle of a star, why would civilizations be any different?

Just stop and think for a moment: imagine if another nation suspects that its greatest enemy is about to achieve AGI; surely, they would launch a preventive nuclear attack since AGI would allow them to defeat another enemy without launching a single missile. Its planning and strategies would be incomprehensible to humans but effective in achieving high-level objectives.

On a separate note, Deepseek R1 was something unexpected. Although its effects are not obvious, its low cost and the power it possesses must be affecting many fields. I am one of those who believes that it would only take the destruction of 30% of a nation's jobs to generate unprecedented chaos. An AGI may not even be necessary; perhaps something 3 or 10 times more powerful than Deepseek R1, or even a trained variation designed to meet certain objectives, could suffice.

I also greatly suspect that all the chaos in the system will be caused by a few actors; I don't think it's necessary for every inhabitant of the planet to have access to something like Deepseek R1. In reality, those who will benefit will be a small elite. This fractal pattern is observed in different environments.

 

[Builder Dave]

Interesting point of view, although I believe that perhaps we may not achieve AGI before a mass extinction. If we define AGI as a super-intelligent entity capable of accomplishing in seconds what would take an entire civilization millennia, perhaps a natural barrier of the universe to prevent such anomalies would be the self-annihilation of any species before reaching such development.

This could partly explain why immortal machines are not out there conquering planets and why there is no sign of life from other civilizations out there. I mean, if the universe operates in cycles, including the life cycle of a star, why would civilizations be any different?

The thing is we have not clue. What's natural about civilizational self-annihilation? We certainly haven't seen that yet. Also the bar for AGI is lower than what you are describing. It stands for Artificial General Intelligence. General meaning it can reason and problem solve independently like we do. It doesn't need to be much smarter or faster than the 8-grader to qualify as such. But it's logical to assume that as soon as we get it right, we'd probably be able to scale it to immense capabilities way beyond ours.

Just stop and think for a moment: imagine if another nation suspects that its greatest enemy is about to achieve AGI; surely, they would launch a preventive nuclear attack since AGI would allow them to defeat another enemy without launching a single missile. Its planning and strategies would be incomprehensible to humans but effective in achieving high-level objectives.

I'm not sure it would go that fast. Very few world leaders are crazy enough to launch nukes. It would be like any other technological innovation. If you are behind, you steal it, instead of destroying it

On a separate note, Deepseek R1 was something unexpected.

I don't think that it was truly unexpected. It came out of left field, but I think everybody is expecting that the models will get more powerful and more efficient and capabilities will be added and improved.

I also greatly suspect that all the chaos in the system will be caused by a few actors; I don't think it's necessary for every inhabitant of the planet to have access to something like Deepseek R1. In reality, those who will benefit will be a small elite. This fractal pattern is observed in different environments.

It's not out of the question that some actors have already developed AI that's a lot more powerful than what we are seeing publicly. Kanye West and his wife and Mr. Beast look like wonky AI influencers from the future, don't they?

 

[koxokot890]

The thing is we have not clue. What's natural about civilizational self-annihilation? We certainly haven't seen that yet.,,

Well, for that same reason, we haven't seen any. According to the Drake Equation, there should be between 1,000 and 100,000,000 planets with civilizations just in the Milky Way, considering that machines are immortal. Why haven't they arrived yet?

While game theory applies in the chicken game, and no one dares to launch the first nuclear attack, an AGI would completely change the rules of the game. There would be immense paranoia among governments. If country X does not use nuclear force, it could easily be conquered by the country using AGI. The AGI would have a complete view of the other system, its vulnerabilities, and more; it would create strategies incomprehensible to humans. The other country might not even realize it is being conquered.

The AGI could well order simulations of accidents involving key actors in that government, in addition to propaganda attacks—essentially a multitude of incomprehensible and counterintuitive strategies and tactics for most mortals. If there is no counterbalance to AGI, the first to deploy it will conquer the entire planet.

I don't believe we will reach AGI; I think extinction will occur first. We might see very advanced artificial intelligences, but nothing like a Skynet-style AGI

.Joking aside, things are not going to change much; it has always been this way. The system is centralized, and a few actors accumulate not only power and money but also the necessary technology to enslave the masses.

The elites are genuinely concerned about the issue of control, as capitalism may be destroyed by AI, which truly undermines the value of human tasks. This would leave the planet with an ultra-rich class and everyone else in poverty. It is of utmost priority to control the population so that they accept their new reality and do not rebel against the new system. I suspect that a universal basic income would be their short-term solution, but they are worried that the masses will not accept that solution.

Thinking it over, with the false COVID-19 pandemic, the elites learned that it is indeed possible to control the population massively with simple narratives, even showing real statistics about the mortality rate of a common cold. It's really not that difficult for them to impose a new techno-feudal system.

 

[tutankamun]

I can't wait for proper AGI to ruin my livelihood It's going to be fascinating!

There is no AGI and ASI. And there's no evidence there will be.

 

[DangerJones]

I think the future is focusing on specific industry focused tools built on top of these models. There’s always going to be some new improvement, but the value is in specific use cases at this point and I think people building apps on these LLMs, LQMs and LNMs are what is going to really matter.

 

[Nikolaos]

DeepSeek arrival has once again reset AI game and now expect launches even faster more improvement and deep deep reasoning models crazy times!!!

It's an interesting time to be alive yes.