Cerebrium

cerebrium

Finding the right AI model to build a workflow around is hard. With so many models available across different platforms, it’s impossible to know where to start or how to find the one that best fits your specific needs. That’s the problem I set out to solve with AIModels.fyi, the only search engine for discovering and comparing AI models from across the web. The site has a simple premise: describe your problem in natural language, and get back a list of AI models that can potentially solve it. For example, you might search for “I need a model that can help upscale my images without losing clarity” or “make my cat look like it’s singing”. The goal is to , not just ML engineers, to find useful AI tools for their projects. make it easy for anyone Under the hood, the search works by taking the user’s query and comparing it to a database of over 240,000 AI models from 7,500+ creators across major platforms like Hugging Face, Replicate, Cerebrium, and DeepInfra. When a user types in a search, the query gets embedded into a vector using OpenAI’s text-embedding-ada-002 model. This query vector is then compared to pre-computed vector representations of the model names, descriptions, and use cases in the database. The most similar models by cosine distance are returned as the search results. Here’s a simplified version of the core search logic I’ve been using for about six months: const fetchData = async (query) => {
  // Embed the search query
  const embeddingResponse = await openAi.createEmbedding({
    model: "text-embedding-ada-002",
    input: query,
  });
  
  const embedding = embeddingResponse.data.data[0].embedding;

// Find the most similar models in the vector database 
  const { data: modelsData, error } = await supabase.rpc("search_models", {
    query_embedding: embedding,
    similarity_threshold: 0.75,
    match_count: 10,
  });
  return modelsData;
}; This vector search approach works reasonably well for surfacing relevant models when there’s a direct overlap between the query terms and the model descriptions. However, it struggles with more abstract, complex, or niche queries where the user’s language doesn’t exactly match the technical terminology in the database. For example, I recently wanted to find a model that could animate a portrait image using an audio voice clip to create a talking head effect. I knew there were some good models out there for this, but my searches for things like “talking portrait from image and audio” weren’t turning up the results I expected. The problem was that my query terms didn’t have enough literal matches with the specific words used in the model descriptions, even though the semantic intent was similar. This got me thinking about ways to improve the search experience to better handle those kinds of abstract or complex queries. How could I make the system smarter at understanding the user’s true intent and mapping it to relevant models in the database? That’s when I came across the idea of Hypothetical Document Embeddings (HyDE) from after reading about it in . this research paper this Medium post Meet Mr. HyDE — The query expander! The key insight behind HyDE is to use a large language model (LLM) to expand the user’s short query into a more detailed hypothetical “ideal” document before embedding it for search. So, instead of directly matching the query embedding to the database embeddings, it first translates the query into a richer representation in the same semantic space as the real documents. Here’s how the process works: The user enters a natural language query expressing their needs, like “I want to convert an image into a 3D model.” This generated pseudo-document is then embedded into a dense vector using a model like OpenAI’s text-embedding-ada-002. The resulting embedding is used to query the vector database and find the most similar real model documents. The query is fed into an instruction-tuned LLM like GPT-3 to generate a hypothetical document. The prompt asks the model to imagine an ideal document for answering the query. For example: Query: I want to convert an image into a 3D model Hypothetical Document: There are several AI models available that can convert a 2D image into a 3D model. The process is known as single-view reconstruction or 3D estimation. It typically involves using deep learning to predict the depth map and 3D shape from the 2D input image. Some popular models for this task include Pix2Vox, Pixel2Mesh, AtlasNet, and OccNet. These models are trained on large datasets of 3D objects and can handle complex shapes. To use them, you provide an input image and the model will output a 3D mesh or voxel representation that you can render or edit further in 3D modeling tools… The LLM step helps bridge the gap between the user’s simple query and the more technical language used in the model descriptions. It expands on the initial query to infer additional relevant terms, concepts, and context that can help retrieve better matches. At the same time, using real document embeddings as the final search index keeps the results grounded and prevents hallucination. I implemented a basic version of HyDE on top of the existing AIModels search, and the initial results are promising. In some sample queries I tested, HyDE surfaced relevant models that the original keyword-matching approach missed. For the talking head portrait query, it was able to generate a hypothetical document with key terms like “lip sync”, “facial animation”, and “audio-driven animation”, which helped retrieve some strong model matches that the plain search had overlooked. two to three times more How to implement HyDE in your own search system Let me show you how HyDE can be added in on top of an existing semantic search system to dramatically improve its quality. Let’s start with a simplified code sketch based on my current search flow. Here’s the full file — you can glance through it in totality before we look at the individual parts: const fetchData = async (query) => {
  // Generate a hypothetical document embedding for the query 
  const hypotheticalDocument = await generateHypotheticalDocument(query);
  
  const embeddingResponse = await openAi.createEmbedding({
    model: "text-embedding-ada-002",
    input: hypotheticalDocument,
  });
  
  const embedding = embeddingResponse.data.data[0].embedding;

// Search the real model embeddings using the hypothetical document embedding
  const { data: modelsData, error } = await supabase.rpc("search_models", {
    query_embedding: embedding,
    similarity_threshold: 0.75,
    match_count: 10,
  });
  return modelsData;
};
const generateHypotheticalDocument = async (query) => {
  // Use GPT-3 to expand the query into a hypothetical ideal document
  const prompt = `The user entered the following search query: 
  "${query}"
  
  Please generate a detailed hypothetical document that would be highly relevant to answering this query. The document should use more technical language and expand on the key aspects of the query.
  Hypothetical Document:`;
  const response = await openAi.createCompletion({
    model: "text-davinci-002",
    prompt: prompt,
    max_tokens: 200,
    n: 1,
    stop: null,
    temperature: 0.5,
  });
  const hypotheticalDocument = response.data.choices[0].text.trim();
  return hypotheticalDocument;
} Let’s see how it works. We’ll start with the function, which is the main entry point for the search process. First up: fetchData const hypotheticalDocument = await generateHypotheticalDocument(query); This line calls the function (which we'll explain in more detail later) to generate a hypothetical ideal document based on the user's search query. generateHypotheticalDocument The hypothetical document is a key part of the HyDE approach, as it aims to capture the user’s search intent in a more detailed and technical way than the original query. const embeddingResponse = await openAi.createEmbedding({
  model: "text-embedding-ada-002",
  input: hypotheticalDocument,
});
const embedding = embeddingResponse.data.data[0].embedding; These lines use OpenAI’s model to generate an embedding vector for the hypothetical document. I think it's the best embedding option out there so far. text-embedding-ada-002 Embeddings are a way to represent text as a dense vector of numbers, where similar texts have similar vectors. This allows us to perform semantic similarity searches. By embedding the hypothetical document instead of the original query, we aim to get a vector representation that better captures the expanded search intent. Next up: const { data: modelsData, error } = await supabase.rpc("search_models", {
  query_embedding: embedding,
  similarity_threshold: 0.75,
  match_count: 10,
}); This part performs the actual similarity search in the model database. It is called a remote procedure ( ) on the Supabase database, passing in the embedding of the hypothetical document as the . search_models query_embedding The and parameters control how closely the results need to match and how many to return. similarity_threshold match_count The search looks for model embeddings that are most similar to the query embedding based on cosine similarity. Now let’s look at the function: generateHypotheticalDocument const prompt = `The user entered the following search query: 
"${query}"

Please generate a detailed hypothetical document that would be highly relevant to answering this query. The document should use more technical language and expand on the key aspects of the query.
Hypothetical Document:`; This part constructs the prompt that will be sent to the language model to generate the hypothetical document. The prompt includes the user’s original query and instructions for the model to generate a detailed, technical document that expands on the query’s key aspects. Crafting an effective prompt is crucial for guiding the language model to generate useful hypothetical documents. Now for the next part: const response = await openAi.createCompletion({
  model: "text-davinci-002",
  prompt: prompt,
  max_tokens: 200,
  n: 1,
  stop: null,
  temperature: 0.5,
}); These lines use OpenAI’s model to generate the hypothetical document based on the prompt. text-davinci-002 The parameter limits the length of the generated document while controls the randomness (higher values make the output more diverse but potentially less focused). Adjusting these parameters can help tune the quality and variety of the generated documents. max_tokens temperature Next, we have: const hypotheticalDocument = response.data.choices[0].text.trim(); This line extracts the generated hypothetical document text from the GPT-3 API response. The generated text is then returned to be used in the embedding step. Then we can just retreive the data as before, this time using the hypothetical document as the query, but getting a lot more relevant results due to the expanded search term and all it’s extra, juicy context! Recap — what did we just build? I hope the code and my explanation demonstrate how the HyDE approach can be implemented on top of an existing semantic search system. It’s not that hard! The key idea is this: By inserting the hypothetical document generation and embedding steps into the search pipeline, we aim to improve the search results by better capturing the user’s intent. To recap, the key parts were: Using a language model to expand the user’s query into a more detailed, technical hypothetical document Embedding the hypothetical document to get a vector representation that aligns with the model database Performing a similarity search between the hypothetical document embedding and the model embeddings to find the most relevant results Of course, this is just a simplified implementation (simplementation?), and there are many details and optimizations to consider in a real implementation. But it illustrates the core flow and components of the HyDE approach. Taking it further Since this is just an initial prototype there’s still a lot of room for experimentation and refinement. Some of the key open questions I’m thinking about: What’s the optimal prompt structure and output length for generating useful hypothetical documents? I’d like to try different prompt templates and output sizes to see how they impact result quality. How can the hypothetical documents be made more interpretable and transparent to users? It could be confusing if users don’t understand why certain terms are being added. I’m considering showing parts of the generated text as a “query interpretation” to provide more insight into the process. To what extent does the result quality depend on the specific LLM and embedding models used? I’m curious to benchmark different model combinations (e.g. GPT-4 vs GPT-3, vs OpenAI embeddings, etc.) to find the best recipe. Contriever How can this approach be efficiently scaled to work over millions of database entries? Inference on large LLMs like GPT-3 is expensive and slow. I’ll need to look into caching, optimizing, or distilling the models to make this viable at scale. I’m excited about the potential for language models to help make search interfaces like AIModels more intuitive and powerful for a wider audience. Bridging the semantic gap between how users naturally express their needs and how AI models are technically described feels like an important piece of the puzzle. Techniques like HyDE point a way toward search experiences that deeply understand the intent and meet users where they are. My long-term goal is to keep refining this interface until anyone can easily find relevant AI models for their needs, no matter their level of technical expertise. I’ll be sharing more details as I continue experimenting with HyDE and other approaches. If you’re curious to learn more about the technical side, I recommend checking out the original paper. And if you have ideas or feedback on the AIModels search experience, I’d love to hear from you! Let me know what you think in the comments. Also published . here

Improve Semantic Search and Enhance AI Model Discovery with Hypothetical Document Embeddings (HyDE)

Startups of The Year 2023 Winners: Europe

LangChain is a groundbreaking framework for building applications powered by large language models (LLMs). The main premise behind LangChain is to facilitate the creation of and agentic applications that can interact intelligently with their environment. is super cost-effective partner in the LangChain ecosystem, offers a suite of versatile LLMs that users can leverage to build advanced applications. This article will give you an overview of LangChain and show you how to get started with a quick example using a DeepInfra model. data-aware DeepInfra LangChain Framework At its core, LangChain is that, when combined, offer a powerful platform for creating . These modules include Models, Prompts, Memory, Indexes, Chains, Agents, and Callbacks. composed of several modules LLM applications These represent the different types of LLMs that can be integrated into the framework. Models: This module manages the interaction with the LLM, including optimization and serialization. Prompts: This pertains to the persistence of state across various calls within an application. Memory: These modules allow the LLM to interact with application-specific data, making it more powerful and context-aware. Indexes: These structured sequences can call an LLM or another utility to perform a specific task. Chains: An agent is essentially a chain in which an LLM, given a high-level directive and a set of tools, can execute actions and observe outcomes until the directive is complete. Agents: **Callbacks:**These enable developers to log and stream the intermediate steps of any chain, facilitating debugging and evaluation. Use Cases of LangChain LangChain, thanks to its flexibility, can be utilized for a wide range of applications, from Autonomous Agents to Evaluation. Here are a few examples of what LangChain can be used for: LangChain can be used to create intelligent assistants that can interact, remember previous interactions, and gain knowledge over time. : Personal Assistants : LangChain can answer specific questions by referring only to designated documents, thus providing accurate and reliable responses. Question Answering LangChain is ideal for creating chatbots thanks to its ability to manage complex conversations. : Chatbots LangChain can be used to analyze and understand code, making it valuable for software development. : Code Understanding LangChain can interact with APIs, thereby providing access to real-time information and enabling action execution. Interacting with APIs: LangChain can compress large amounts of data and provide concise summaries. Summarization: Integration of DeepInfra's LLMs into LangChain Integrating into LangChain is straightforward, as outlined below: DeepInfra's LLMs Set the Environment API Key Obtain your API key from and set it as an environment variable. DeepInfra Create the DeepInfra instance Choose your model from DeepInfra and set model-specific parameters. For instance, you can use model to create this simple demo app: databricks/dolly-v2-12b from langchain.llms import DeepInfra
llm = DeepInfra(model_id="databricks/dolly-v2-12b")
llm.model_kwargs =
{'temperature': 0.7, 'repetition_penalty': 1.2, 'max_new_tokens': 250, 'top_p': 0.9} Create a Prompt Template Create a template for your question and answer. from langchain import PromptTemplate

    template = """Question: {question}\n\nAnswer: Let's think step by step."""
    prompt = PromptTemplate(template=template, input_variables=["question"]) Initiate the LLMChain Prepare the LLMChain with your prompt and LLM. from langchain import LLMChain
llm_chain = LLMChain(prompt=prompt, llm=llm) Run the LLMChain: Run the LLMChain with a sample question and get the response. question = "Can penguins reach the North pole?"
print(llm_chain.run(question)) LangChain Ecosystem The LangChain ecosystem is a vibrant and dynamic space that provides you with a multitude of opportunities. It connects you with a diverse range of Large Language Model (LLM) providers and AI-focused services, expanding the capabilities and possibilities of your LangChain applications. You can integrate with leading LLM providers such as AI21, Aleph Alpha, and Databricks, among others, to leverage their advanced models. Additionally, you can tap into the power of AI platforms like Google's Vertex AI and Amazon's SageMaker to enhance your projects further. LangChain also seamlessly synergizes with specialized toolkits like Hugging Face for efficient local pipeline management, ensuring smooth and streamlined workflows. With JSONFormer, you can easily decode and structure the output of LLMs, making it easier to work with the generated data. Moreover, LangChain offers extensive infrastructure support through tools like Beam for parallel data processing and Runhouse for efficient environment management. This wide network of integrations enriches the LangChain environment, empowering you as a developer to create sophisticated and feature-rich applications with relative ease. You have the resources and support needed to unlock the full potential of LangChain and build innovative solutions that harness the power of AI. Deployment and Reference Docs LangChain provides you with a wealth of resources to guide you from the initial installation phase to the final deployment of your application. Comprehensive documentation is at your disposal, offering detailed instructions on integrating and deploying LangChain applications. This helps smooth your transition from development to production and addresses potential obstacles you might encounter. One of the key resources is , a powerful tool that enables you to search, filter, and sort through hundreds of AI models, including a wide selection from DeepInfra. This platform is designed to help you find the right model for your specific AI project. Plus, you can subscribe for monthly updates on new models, keeping you informed of the latest advancements in the field. AIModels.fyi The site also offers tools that allow you to compare models across various platforms, such as Replicate and Cerebrium. This means you can make more informed decisions when selecting a model based on your project's requirements and the model's capabilities. With LangChain's extensive documentation, robust resources, and comparative tools, you're well-equipped to maximize the benefits of LangChain in your projects. Conclusion In summary, LangChain, coupled with DeepInfra's LLMs, represents a powerful tool for developers aiming to create sophisticated language model applications. The flexibility and versatility of LangChain make it an ideal choice for a variety of use cases. I invite you to explore and leverage LangChain, especially in combination with DeepInfra's LLMs, to create innovative applications that can truly harness the power of AI. Also published . here

An Introduction to LangChain Use Cases with DeepInfra

There's been a lot of fuss about the dangers lurking within the advancement in tech, and the impact it can have in various fields across the human-job-sphere. Well, evidently so due to the recent loss of jobs for 'manual workers' or carefully put, displacement of most manual workers with machines. However, we can not ignore the formidable synergy from the man/machine merger that has pierced into uncharted historical territories, pushing the frontier of innovation, imagination, and human possibilities. This synergy is the birthplace of almost every breathtaking innovation. Among several realms of scientific innovations being birthed, only a few fields hold as much intrigue and promise as neuroprosthetics. Neuroprosthetics can be seen as one of the perfect scenarios where technology heavily overlaps medicine as it pushes the bandwidth of possibilities even further. Stephen Hawking once opined, We are all now connected by the Internet, like neurons in a giant brain But with the advent and advancement of neuroprosthetics, it is evident that this connection is becoming more literal than metaphorical. What is Neuroprosthetics? Neuroprosthetics is an intersectoral field of neuroscience and engineering that employs technological innovation with scientific baselines to develop devices that can restore the lost function of the nervous system. These devices called Neuroprostheses can restore or modulate varieties of neurally-mediated functions for the treatment of a range of conditions caused by neurodegenerative diseases, congenital anomalies, infections, and accidents. These conditions include spinal cord injury, Parkinson's disease, stroke, blindness, deafness, and others caused by accidents and attacks on the battlefield. History of Neuroprosthetics We can not categorically pinpoint the genesis of neuroprosthetics but history has a record of ancient Egyptians using leather and wood to construct artificial limbs as far back as 950 B.C. Several other prostheses like the Cairo Toe (700-950 B.C.), the Cartonnage Toe ( 600 B.C. ), and the Roman artificial leg (300 B.C. ) were discovered in ancient tombs and graveyards. In retrospect, it can be observed that ancient scientific knowledge was limited to artificial limb construction. However, this was a building block for a more sophisticated system as presented today. To understand how we got to this level of sophistication, it is pertinent to lean back to memory lane and understand the chronological progression of neuroprosthetics. The following are some major historical strides in neuroprosthetics: 1579 - Ambroise Paré designed prosthetic limbs with mechanical components 1791 - Lurgi Galvani discovered that electrical stimulation of the nerve can use muscle contraction. This discovery was a huge milestone in the study of the nervous system, laying the framework for the creation of neuroprosthetics. 1860s - Gustav Fritsch and Eduard Hitzig via an experiment discovered that electrical stimulation of a certain part of the cerebrum causes involuntary muscular contractions of specific parts of the body. This discovery paved the way for advancement in BCIs. 1950s - José Manuel Rodriguez Delgado orchestrated the brain electrode implantation that uses electricity to stimulate specific brain areas for the treatment of epilepsy and other types of mental conditions. 1957 - The creation of the first cochlear implant as a hearing aid. 1960 - Testing and subsequent commercialization of neuroprosthesis for the treatment of drop-foot. 1977 - The development of the first Auditory Brainstem Implant (ABI) for deaf patients who are not eligible for cochlear implants. 1981 - Implantation of a peripheral nerve bridge in the spinal cord of a rat as a conduit for regenerating nerve fibers. 1988 - Development of the lumbar anterior root implant and Functional Electrical Stimulation (FES) for the treatment of paraplegia. 1990 - Approval of the first Cochlear implant in children by FDA. 1999 - John Chapin and his team experimented to demonstrate the feasibility of using brain activity for real-time control by simultaneous neuronal population recordings in experimental rats. 2010s - The development of close-looped neuroprosthetics that allowed for more effective and precise control of neuroprosthesis. The 21st century has undoubtedly witnessed astronomical-paced development in neuroprosthetics, as there have been several remarkable advancements and sophistication of BCIs and other neuroprosthetics. For instance; The development of Brain-controlled robotic limbs for individuals with limb amputations. The design of retinal implants and visual neuroprosthetics to restore vision. The ongoing development of neurotrophic factors which are proteins predicted to be a potent treatment of spinal cord injuries and other neurological disorders by facilitating and enhancing the growth of neurons. The ongoing development of a novel type of BCI called Neural Dust anchored by Elon Musk’s Neuralink. Classes of Neuroprotheses Neuroprotheses are classified into two classes which include implantable and non-invasive neuroprotheses. These devices are surgically buried in the body for direct interaction with the spinal cord, brain, peripheral nerves, or at times the entire nervous system. The design of these devices enables them to either record neural activities in the nervous system, stimulate neurons, or foster communication between neural circuits and external devices. Examples include brain stimulators, cochlear implants, retinal prostheses, and others. Implantable Neuroprotheses: These are brain-computer interface and myoelectric prostheses which are worn on the body and as such do not have any direct interaction with the nervous system. Non-Invasive Neuroprotheses: Types of Neuroprostheses All types of neuroprostheses share a similar mode of operation irrespective of their type. They create a neural link between the brain and a computer, enabling a patient to use his/her nervous system to control sensory organs or limbs. The primary types of neuroprostheses include the sensory, motor, and neuromodulation neuroprostheses. Sensory Neuroprostheses These types of devices convert light, sound, and other physical stimuli to electrical signals that the brain can interpret to enhance lost sensory modalities. Examples of Sensory neuroprostheses include: These devices stimulate the photoreceptor cells in the degenerated retinal cells to send visual information to the brain. Example: the retinal implants. Visual Prostheses: These devices bypass the damaged cochlea and directly stimulate the auditory nerves to enable the patient to hear sounds. Example: Cochlear implants Auditory Prostheses: Motor Neuroprostheses These devices interpret the patient’s intended movements via signals from the muscle or brain to act. Examples of Motor prostheses include: Controlled by the myoelectric signals, prosthetic limbs are artificial limbs that replace missing or amputated body parts to provide mobility and support. They are mechanical devices that get electrical signals from the user’s muscles via the myoelectric signals to control movements. Prosthetic Limbs: These are electrodes that utilize electrical currents to stimulate muscles and nerves directly to enhance or restore the functions of the muscles and nerves marred by neurological conditions or perhaps paralysis. These electrodes are placed externally on the skin or implanted surgically into the skin. Paralytic patients on FES devices can move their limbs with ease. Functional Electrical Simulation (FES) Devices: These are implantable neuroprostheses that use brain signals to enable users to control external devices like wheelchairs or robotic arms. These devices work by capturing and analyzing the electrical activities of the brain’s neurons and utilizing this information to control external devices. can also be non-invasive like Electroencephalography (EEG) which is used to record signals from the scalp. Brain-Computer Interfaces (BCIs): BCI Neuromodulation Neuroprosthetics These devices use electrical signals to regulate and modulate the activities of the nervous system to treat a plethora of abnormalities like neurological diseases, movement disorders, and diseases. Examples of these devices include; These are brain implantable devices that deliver electrical stimulation to specific areas to modulate neural activities in these areas. They are used to treat a variety of conditions including dystonia, Parkinson’s disease, essential tremor, obsessive-compulsive disorder (OCD), major depressive disorder (MDD), and Tourette syndrome. Deep Brain Stimulators ( ): DBS They are non-invasive neuromodulation devices that use strong magnetic fields to stimulate brain nerve cells. They treat medical conditions such as depression, anxiety, PTSD, OCD, and migraines. Transcranial Magnetic Stimulators (TMS): This device mimics the spinal cord by modulating the signal transmitted between the brain and the body and sending the same to the spinal cords to aid people with spinal injuries and to treat chronic pains like neuropathic pains and back pains. Spinal Cord Stimulator: These devices are implanted in the vagus nerve to stimulate this nerve to treat conditions such as epilepsy, depression, and migraine. The Vagus nerve connects the brain to the rest of the body and plays an active role in regulating various functions in the body. Vagus Nerve Stimulators (VNS): These devices are implanted next to one of the peripheral nerves. They function by using electric currents to alter the central processing of pain signals for the treatment of total knee arthroplasty, chronic knee pains, and back pain. Peripheral Nerve Stimulators (PMS): These devices establish bi-directional communication (that is: a real-time sensing and feedback mechanism) with the nervous system to enable a deeper level of interaction for a more precise neuromodulation. An example is a Responsive Neurostimulation (RNS) Device used on epileptic patients to prevent seizures. Closed-loop neuroprosthetics: Ethical Consideration in Neuroprosthetics Neurodiversity is defined as the variation in neural and mental functions arising naturally from the interactions between genetic, neurobiological, and environmental factors. These functions are believed to exist on a spectrum that ranges from the normal to the abnormal. While some regard functions in the abnormal axis as disabilities, others choose to perceive them as identities that should be recognized, respected, and preserved. This viewpoint is premised on the fact that a person with abnormal (pathological, to be precise) neural sense has created a world around the abnormality. Besides, it is believed that the absence of a sensory capability triggers a supernatural functionality of one or more sensory capabilities to compensate for this loss. In an instance where the lost sensory capability is restored, it is difficult to adequately predict the effect this will have on the already enhanced capacities. Moreover, the radical psychological and lifestyle adjustment necessary for the integration into a new experiential world might be difficult as the patient has to reshape his perception of the world and other realities he was previously denied. This is why critics argue that to measure the entire worth of the prosthesis, psychological, social, and cultural elements related to a sensory capability must be examined alongside the neurophysiological ability to use that sense organ. Although neurodiversity is usually mentioned in the debate on the use of neuroprosthetics on patients with autism spectrum disorder to modulate neural circuits, its core values and concepts can be stretched to cover other neurological and psychiatric conditions. For conditions such as bipolar or unipolar, patients may choose to resist treatment and identify with the cognitive and affective states of the disorder. This identification is believed to have influenced legal reasons against forcibly using neuroprosthetics or psychotropic drugs to treat symptoms of these disorders. For neuromodulating prosthetics like DBS, BCI, and Hippocampal prosthetics, controlling symptoms is not the only assessment metric to quantify its effectiveness. Other factors such as the patient's overall experience after using the prosthetics are also considered. In using DBS to treat psychiatric disorders, there have been ethical concerns and questions such as; Did the person receiving the treatment give consent to undergo the procedure? During the modulation of an unconscious patient’s mental function, who is the supreme controller of the patient’s action? The patient or the machine? Is the patient still acting in his free will? How detrimental or beneficial are the neural changes induced by DBS? What is the net gain or net loss? In the case of using BCIs to treat a variety of neurological disorders, which is basically by predicting the intended action of the patients, several authors and researchers have expressed concerns about behavior control. The questions border on the free will of the patients and the algorithm’s functionality to follow the patient’s intentions to the latter. Another general concern stems from the fear that this level of development prevalent in the neuroprosthetic field catalyzed by the advancement in AI, will birth ultra-sophisticated neuroprosthetics that will replace the neural circuit in the brain, making the brain a completely artificial organ. The implication of this can not be overestimated as humans can easily be turned into cyborgs, a stage no one wants to get to. Therefore while neuroprosthetics is gradually advancing its capabilities to completely eradicate neural disorders, there is a need to develop and implement stringent measures of laid-down principles and ethical guidelines to streamline and curb its possibilities of impending doom. End Notes For the past 5 centuries, neuroprosthetics has experienced tremendous evolution from being a practice revolving around a set of crude and rudimentary mechanical devices to highly advanced technologies and innovations that directly interfere with the nervous system. These advancements have improved the quality of life for individuals with disabilities and continue to hold promise for the future, as interdisciplinary research, deep work, and technological innovations propel the field forward. Though not without challenges, this field of science and technology has shown intriguing possibilities for a new reality.

Neuroprosthetics: The Journey So Far

ON THE CARE OF HEALTH.

THE SCIENCE OF INSTINCT

MATHEMATICAL MEMORIES: NEWTON'S BINOMIAL THEOREM

portrays a grim future in which people’s minds are trapped in a computer-generated simulation while their bodies are harvested for electricity to power robotic metropolises. The Matrix Similarly, modern society’s dependence on technology is not only tangible — our reliance on mechanized means of production for our sustenance and standards of living — but also intangible. Indeed, our growing dependence on digital mediums for subsistence and leisure makes the dystopian future portrayed in bear an increasing resemblance to our reality. The Matrix The deepening human-machine dependence From electricity and refrigeration to transportation and medicine, modern standards of living depend on machines. Similarly, in , the residents of Zion — humankind’s last surviving colony on what remains of Earth after sentient robots usurped control — rely heavily on machinery to survive. As Hamann, a senior council member, notes during a conversation with Neo, “I like to be reminded that this city survives because of these machines.”¹ He gestures toward the vast underground apparatus recycling Zion’s air and water supply, “I think about all those people still plugged into the Matrix and when I look at these machines, I can’t help thinking that in a way, we are plugged into them.”² The Matrix Neo challenges Hamann’s characterization of the human-machine relationship, “We control these machines, they don’t control us…If we wanted, we could shut [them] down.”³ While it is true that the machines keeping Zion alive are created and operated by humans and are no more sentient than our kitchen appliances, turning them off is not a viable option. Zion’s survival depends on the machinery just as much as the machinery depends on Zion for its continued operation. The parallels between the survivors living in Zion and modern society are undeniable but hardly surprising: they both share the same basic needs and necessities such as air to breathe, water to drink, and food to eat. More striking are the parallels between our reality and the Matrix-conjured virtual world and its inhabitants. From online dating to remote work, success in our personal and professional lives increasingly hinges on what happens on-screen. As the boundary between the digital and analog worlds becomes increasingly blurred, these similarities are growing stronger. We have come to rely on the Internet for navigation, banking, communication, and more. Moreover, from online dating to remote work, success in our personal and professional lives increasingly hinges on what happens on-screen. Even our idle time is spent online: streaming shows on Netflix, watching short-form videos on TikTok, listening to music on Spotify, and countless other forms of digital entertainment. Leisure time used to be active, involving physical interactions with others or with our environment — in the ‘real’ world. Now, however, much of our interactions are digital, taking place in a virtual world. According to a study by Nielsen, the average American adult spent almost 12 hours on-screen in 2019⁴ — nearly three-quarters of their waking hours. More recent studies suggest that screen time has continued to increase around the world, partly due to the lasting effects of the Covid-19 lockdowns. As a result, we have become more passive — a culture of readers rather than speakers, spectators rather than direct participants. This shift is a natural consequence of having much of what we do, think, and feel transpire via a digital medium, such as our smartphones. We might as well be plugged into those devices like the unwitting masses in . The Matrix More than science fiction The parallels between our reality and Wachowskis’ imagination are more than figurative. Every product catering to our digital cravings runs on vast data centers, each consuming as much electricity as a small town. In the United States alone, data centers account for approximately 2% of the country’s total electricity usage,⁵ equivalent to the energy consumption of almost 7 million people.⁶ It is the users’ time and attention — hours of their lives every day — that pays for the data centers’ maintenance and energy bills. We call them “server farms,” but it is the users whose time and data are harvested, making screen-glued netizens eerily reminiscent of the Matrix-plugged souls serving as robotic batteries. The human-machine dependence is mutual — in the movie and the real world alike. Though subjugated, humankind in becomes the lifeblood of the robotic world. Machines grow, harvest, and recycle human bodies in vast dystopian farms as their sole source of electricity. Humans are to machines as agriculture is to humankind. Agent Smith, a sentient program designed by the Machines to keep order in the Matrix, even refers to humans as “crops.”⁷ The Matrix We call them “server farms,” but it is the users whose time and data are harvested, making screen-glued netizens eerily reminiscent of the Matrix-plugged souls serving as robotic batteries. Modern society’s digital pastimes have a similar dependence on their users. The more time we spend interacting with digital entertainment products, the better they get at catering to our cravings. Netflix learns to recommend the types of shows we prefer; TikTok fine-tunes its videos to match our fancy; Spotify adapts its playlists to mirror our music taste. The data we produce through our interactions is the fodder for the recommendation algorithms powering these platforms. The user base of digital platforms is vital not only to their proper functioning but also to their very survival. There would be no or without subscribers and no TikTok or Instagram without the ad revenue made possible by their vast audiences. The history of the Internet, however brief, is littered with corpses of former social media giants that suffocated as their user base dwindled: MySpace, Friendster, Vine, and many more met such untimely demise. Those that survived, such as Yahoo or AOL, are mere shadows of their former selves. Netflix Spotify The Machines operating the Matrix are comparably reliant on their ‘userbase.’ When human crop yields declined in the wake of the ceasefire Neo made possible, it caused an energy shortage so large that Machines turned on each other. General Niobe, humankind’s new leader, retells the events following the human-machine truce: “Nothing can breed violence like scarcity. For the first time, we saw Machines at war with one another.”⁸ The strife for survival in the face of scarce resources echoes the fierce competition among digital giants for our time and attention. Red pill or blue pill? To maximize user engagement, the Internet is shaped by our desire for instant gratification and sensationalism over meaningful, in-depth content. It is a mirror of our desires, not aspirations. Sensationalist, clickbait headlines attract more eyeballs than balanced, long-form journalism. Users’ attention is drawn to tabloid, tragedy, and mindless content such as cat or dance videos in an endless dopamine-driven feedback loop. Similarly, the Matrix is created not as an ideal world — free of war, famine, and disease — but as one its inhabitants desired, with all the good and the bad that comes with the human condition. Agent Smith observes, “I believe that, as a species, human beings define their reality through suffering and misery. The perfect world was a dream that your primitive cerebrum kept trying to wake up from.”⁹ Comparing Smith’s observation to the kind of online content that draws the most views and the highest engagement, it is clear that it applies equally to the present-day Internet. It is a reminder that we must be mindful of the content we consume and the role it plays in shaping not only our online spaces but also our perception of reality and each other. Unlike the denizens of the Matrix, who are connected to the digital simulation unknowingly and without consent, we have a great deal of agency over how we spend our time. No one forces us to download TikTok or watch Netflix; we do so by choice. We choose to engage in passive consumption, often over reconnecting with old friends, going to the gym, or doing any number of other activities we, deep down, know we should be doing instead. In the near future, technologies under active development today may bring humanity ever closer to . Mark Zuckerberg, Meta’s chief executive, makes no secret of his company’s ambition to create a ‘metaverse’ where many of our real-world interactions would be made possible through virtual reality with ever-increasing fidelity.¹⁰ That we would enter those worlds by donning a virtual reality headset rather than by plugging a cable into our nape like Neo or Trinity would become an increasingly meaningless distinction. The Matrix As technology continues to penetrate ever deeper into our lives, we must reflect on our digital habits and consider the implications of our increasing dependence on our devices. Screen time should enhance our lives, not hinder them. Like Neo confronted by Morpheus in ’s most iconic scene — a colorful pill in each hand — the choice is ultimately ours. The Matrix Lead image generated with stable diffusion. Prompt: Illustrate characters from the Matrix This article first appeared here. References [1] (2003), directed by Lana Wachowski & Lilly Wachowski. The Matrix Reloaded [2] Ibid. [3] Ibid. [4] (2019), The Nielsen Company. The Nielsen Total Audience Report [5] (2023), Office of Energy Efficiency & Renewable Energy. Data Centers and Servers [6] Based on compiled by the . statistics U.S. Department of Energy [7] (1999), directed by Lana Wachowski & Lilly Wachowski. The Matrix [8] (2021), directed by Lana Wachowski. Matrix Resurrections [9] (1999), directed by Lana Wachowski & Lilly Wachowski. The Matrix [10] K. Wagner, “ ” (2021), . Zuckerberg Says Facebook’s Future Lies in Virtual Metaverse Bloomberg

From Sci-Fi to Reality: The Matrix as a Mirror of Our Screen-Centric World

সাই-ফাই থেকে বাস্তবে: আমাদের স্ক্রিন-কেন্দ্রিক বিশ্বের আয়না হিসাবে ম্যাট্রিক্স

Von Science-Fiction zur Realität: Die Matrix als Spiegel unserer bildschirmzentrierten Welt

De la ciencia ficción a la realidad: Matrix como espejo de nuestro mundo centrado en la pantalla

De la science-fiction à la réalité : la matrice comme miroir de notre monde centré sur l'écran

विज्ञान-कथा से वास्तविकता तक: हमारी स्क्रीन-केंद्रित दुनिया के दर्पण के रूप में मैट्रिक्स

SF から現実へ: スクリーン中心の世界の鏡としてのマトリックス

공상과학에서 현실까지: 화면 중심 세계를 반영하는 매트릭스

Da ficção científica à realidade: Matrix como um espelho de nosso mundo centrado na tela

Bilim Kurgudan Gerçeğe: Ekran Merkezli Dünyamızın Aynası Olarak Matrix

Từ khoa học viễn tưởng đến hiện thực: Ma trận như một tấm gương phản chiếu thế giới lấy màn hình làm trung tâm của chúng ta

从科幻到现实：矩阵作为我们以屏幕为中心的世界的镜子

A General Introduction to Psychoanalysis by Sigmund Freud, is part of the HackerNoon Books Series. You can jump to any chapter in this book here . PREFACE PREFACE Few, especially in this country, realize that while Freudian themes have rarely found a place on the programs of the American Psychological Association, they have attracted great and growing attention and found frequent elaboration by students of literature, history, biography, sociology, morals and aesthetics, anthropology, education, and religion. They have given the world a new conception of both infancy and adolescence, and shed much new light upon characterology; given us a new and clearer view of sleep, dreams, reveries, and revealed hitherto unknown mental mechanisms common to normal and pathological states and processes, showing that the law of causation extends to the most incoherent acts and even verbigerations in insanity; gone far to clear up the terra incognita of hysteria; taught us to recognize morbid symptoms, often neurotic and psychotic in their germ; revealed the operations of the primitive mind so overlaid and repressed that we had almost lost sight of them; fashioned and used the key of symbolism to unlock many mysticisms of the past; and in addition to all this, affected thousands of cures, established a new prophylaxis, and suggested new tests for character, disposition, and ability, in all combining the practical and theoretic to a degree salutary as it is rare. These twenty-eight lectures to laymen are elementary and almost conversational. Freud sets forth with a frankness almost startling the difficulties and limitations of psychoanalysis, and also describes its main methods and results as only a master and originator of a new school of thought can do. These discourses are at the same time simple and almost confidential, and they trace and sum up the results of thirty years of devoted and painstaking research. While they are not at all controversial, we incidentally see in a clearer light the distinctions between the master and some of his distinguished pupils. A text like this is the most opportune and will naturally more or less supersede all other introductions to the general subject of psychoanalysis. It presents the author in a new light, as an effective and successful popularizer, and is certain to be welcomed not only by the large and growing number of students of psychoanalysis in this country but by the yet larger number of those who wish to begin its study here and elsewhere. The impartial student of Sigmund Freud need not agree with all his conclusions, and indeed, like the present writer, may be unable to make sex so all-dominating a factor in the psychic life of the past and present as Freud deems it to be, to recognize the fact that he is the most original and creative mind in psychology of our generation. Despite the frightful handicap of the odium sexicum, far more formidable today than the odium theologicum, involving as it has done for him lack of academic recognition and even more or less social ostracism, his views have attracted and inspired a brilliant group of minds not only in psychiatry but in many other fields, who have altogether given the world of culture more new and pregnant appercus than those which have come from any other source within the wide domain of humanism. A former student and disciple of Wundt, who recognizes to the full his inestimable services to our science, cannot avoid making certain comparisons. Wundt has had for decades the prestige of a most advantageous academic chair. He founded the first laboratory for experimental psychology, which attracted many of the most gifted and mature students from all lands. By his development of the doctrine of apperception he took psychology forever beyond the old associationism which had ceased to be fruitful. He also established the independence of psychology from physiology, and by his encyclopedic and always thronged lectures, to say nothing of his more or less esoteric seminary, he materially advanced every branch of mental science and extended its influence over the whole wide domain of folklore, mores, language, and primitive religion. His best texts will long constitute a thesaurus which every psychologist must know. Again, like Freud, he inspired students who went beyond him (the Wurzburgers and introspectionists) whose method and results he could not follow. His limitations have grown more and more manifest. He has little use for the unconscious or the abnormal, and for the most part he has lived and wrought in a preevolutionary age and always and everywhere underestimated the genetic standpoint. He never transcends the conventional limits in dealing, as he so rarely does, with sex. Nor does he contribute much likely to be of permanent value in any part of the wide domain of affectivity. We cannot forbear to express the hope that Freud will not repeat Wundt's error in making too abrupt a break with his more advanced pupils like Adler or the Zurich group. It is rather precisely just the topics that Wundt neglects that Freud makes his chief corner-stones, viz., the unconscious, the abnormal, sex, and affectivity generally, with many genetic, especially ontogenetic, but also phylogenetic factors. The Wundtian influence has been great in the past, while Freud has a great present and a yet greater future. In one thing Freud agrees with the introspectionists, viz., in deliberately neglecting the "physiological factor" and building on purely psychological foundations, although for Freud psychology is mainly unconscious, while for the introspectionists it is pure consciousness. Neither he nor his disciples have yet recognized the aid proffered them by students of the autonomic system or by the distinctions between the epicritic and protopathic functions and organs of the cerebrum, although these will doubtless come to have their due place as we know more of the nature and processes of the unconscious mind. If psychologists of the normal have hitherto been too little disposed to recognize the precious contributions to psychology made by the cruel experiments of Nature in mental diseases, we think that the psychoanalysts, who work predominantly in this field, have been somewhat too ready to apply their findings to the operations of the normal mind; but we are optomistic enough to believe that in the end both these errors will vanish and that in the great synthesis of the future that now seems to impend our science will be made vastly richer and deeper on the theoretical side and also far more practical than it has ever been before. G. STANLEY HALL. Clark University, April, 1920. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. Sigmund Freud (2011). A General Introduction to Psychoanalysis. Urbana, Illinois: Project Gutenberg. Retrieved October 2022 https://www.gutenberg.org/cache/epub/38219/pg38219-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org , located at https://www.gutenberg.org/policy/license.html .

PREFACE

A Scientist Discovers a Dimensionless Universe

Stories, News & Mentions about CEREBRIUM

Cerebrium’s Stories on HackerNoon