#DEEPSEEK 🇨🇳 #AI

#AI makes human cognition “Atrophied and Unprepared” #artificialintelligence

Stacker News

Microsoft Study Finds AI Makes Human Cognition “Atrophied and Unprepared” \ stacker news

@co574 stacked 127 sats posting https://www.microsoft.com/en-us/research/uploads/prod/2025/01/lee_2025_ai_critical_thinking_survey.pdf

#AI LLM models predict the next letter by using their knowledge of language and context to make an educated guess! #ArtificialIntelligence #ELI5 Imagine you have a big box of letters and words that you've learned from reading a lot of books, articles, and conversations. When you want to find the next letter in a sentence, the language model looks at the letters and words that have come before it. Here's how it works in simple steps: 1. Understanding Context: The model looks at the words and letters already in the sentence to understand what makes sense next. For example, if the sentence starts with "The cat is," it knows that the next word might be something like "sleeping" or "playing." 2. Probability: The model has learned from all the text it has seen which letters and words are likely to come next. It uses this knowledge to guess the most probable next letter or word. 3. Choosing the Best Option: After considering all the possibilities, the model picks the letter or word that it thinks fits best based on what it has learned.

Deep Sick… How did the Chinese #DeepSeek build a model so good, so quickly and so cheaply; do they know something American AI labs are missing?

DeepSeek Is Chinese But Its AI Models Are From Another Planet

OpenAI and the US are in deep trouble

#AI #ArtificialIntelligence

These two outputs for the prompt “Will Smith eating spaghetti” are just one year apart. At this rate, by the end of the year we’ll have feature-length #AI movies indistinguishable from traditional ones. #ArtificialIntelligence

#AI and the end to all things. #ArtificialIntelligence #longread

Stacker News

AI and the end to all things. \ stacker news

@co574 stacked 53 sats posting https://www.mindprison.cc/p/ai-and-the-end-to-all-things

”…the ocean of behavioral data (data about what we do), in the coming years will feed straight into #artificialintelligence systems. And these systems will remain, to human eyes black boxes. Throughout this process, we will rarely learn about the ‘tribes’ (a statistical group) we belong to or why we belong there. In the era of machine intelligence, most of the variables will remain a mystery” #AI #privacy

#Nvidia named the #Blackwell architecture after David Harold Blackwell, a mathematician who specialized in game theory and statistics and was the first Black scholar inducted into the National Academy of Sciences. #ai #tech

they will try really hard to to rewrite #history with #AI propaganda.

the best #AI directors have been around for less than 3 months…

By the end of 2024, machines will write more than 50% of the new code generated globally. #predictions #ai

When will the first general AI system be devised, tested, and publicly announced? https://www.metaculus.com/questions/5121/ #AGI #AI#metaculus #predictions #forecasting

#AI, the great equalizer? #Artificialintelligence (AI) has often been hailed as having the potential to be the great equalizer - a technology that could help alleviate global inequalities. However, when we examine the development and deployment of AI systems through the lens of complicated, emergent systems, it becomes clear that AI will likely follow the tendency of new technologies to exacerbate rather than ameliorate inequality. A core reason AI will not necessarily be an equalizing force is what we might call "the problem of unequal data." The datasets used to train AI systems reflect wider socio-economic disparities and biases. Facial recognition systems, for example, have been shown to work less accurately for non-white populations. As long as the underlying data reflects existing inequalities, AI will likely propagate and amplify those divides. Treating AI as an impartial, egalitarian technology ignores how it emerges from and interacts with complex human systems and institutions. Relatedly, while AI could improve efficiency and access to services like healthcare for disadvantaged groups, current development trajectories suggest the benefits will likely accrue disproportionately to the already privileged. Those with more resources can dedicate them to harnessing AI, attracting talent to develop new systems towards their own ends. Rather than mitigating inequality, rewards follow the tendency of capitalist markets to centralize resources within dominant groups. Even public goods like healthcare risk optimizing more for those already well-served by the system. Finally, increasing reliance on algorithmic decision-making shifts power into technical systems that few understand and have oversight over. This creates potential vulnerabilities and blindspots - dynamics that historically have harmed marginalized communities most. It remains an open question whether governance mechanisms can be established ensuring AI systems answer equally across populations. The dream that technological progress leads inevitably to greater equality is a stubborn one. But if we have learned anything about complex emergent systems, it is that technology alone does not correct entrenched societal issues—and often exacerbates them. AI is no exception; technological change depends fundamentally on how human values and institutions guide its development. True progress requires conscientious steering of that process towards justice and empowerment for all people. AI will do little to challenge status quos if we do not intentionally shape it to do so.

Uncensored #AI 🛡️ vs #OpenAI 😈

Strategic thinking in #AI should not follow predetermined paths but embrace the chaotic dynamics of evolution. #artificialintelligence #strategy

2026 will be the year of embodied #AI.

The #AI arms race is on!

#AI Agents That Can Bank Themselves. https://mirror.xyz/0x16de9a0d10EFc67EF575e72E40FD4a2A826fdEA7/yBj-fWQxrueXNs8jEjUo-Ozu_Nwt04GaXf1D8LjemtM

The first major attempt at establishing a true field of #artificialintelligence was the Dartmouth workshop in 1956. This would see some of the foremost minds in the fields of mathematics, neuroscience, and computer sciences come together to essentially brainstorm on a way to create what they would term ‘artificial intelligence’, following the more common names at the time like ‘thinking machines’ and automata theory. Despite the hopeful attitude during the 1950s and 1960s, it was soon acknowledged that Artificial Intelligence was a much harder problem than initially assumed. Today, #AI capable of thinking like a human is referred to as artificial general intelligence (#AGI) and still firmly the realm of science-fiction. Much of what we call ‘AI’ today is in fact artificial narrow intelligence (#ANI, or Narrow AI) and encompasses technologies that approach aspects of AGI, but which are generally very limited in their scope and application. Most ANIs are based around artificial neural networks (ANNs) which roughly copy the concepts behind biological neural networks such as those found in the neocortex of mammals, albeit with major differences and simplifications. ANNs like classical NNs and recurrent NNs (RNNs) — what’s used for #chatGPT and Codex — are programmed during training using backpropagation, which is a process that has no biological analog. Essentially, #RNN-based models like chatGPT are curve fitting models, which use regression analysis in order to match a given input with its internal data points, the latter of which are encoded in the weights assigned to the connections within its network. This makes NNs at their core mathematical models, capable of efficiently finding probable matches within their network of parameters. When it comes to chatGPT and similar natural language synthesis systems, their output is therefore based on probability rather than understanding. Therefore much like with any ANN the quality of this output is is highly dependent on the training data set.

The Regulatory Capture of #ArtificialIntelligence by Big Tech is pushing #AI underground.

ExtremeTech

AI Company Plans to Run Clusters of 10,000 Nvidia H100 GPUs in International Waters

Del Complex hopes floating its computer clusters in the middle of the ocean will allow it a whole new level of autonomy—that is, if the company i...