Kozminski Techblog

Reading Time: 4 minutes

Introduction:

Neuralink is an ambitious startup founded by Elon Musk in 2016 that is developing ultra high-bandwidth brain-machine interfaces to connect humans and computers. The goal of Neuralink is to allow humans to directly interface with computers and artificial intelligence through implanted brain chips. Neuralink’s cutting-edge technology involves the development of super-thin threads, measuring less than 10 microns in thickness—smaller than a strand of human hair—that can be implanted into the brain to transmit extensive data. These brain chips facilitate high-bandwidth synaptic transmission and reception of neural signals, paving the way for “full bandwidth data streaming” between the human brain and external devices. 

Aims & Potentials:

The potential applications of this innovation are vast, ranging from restoring sensory and motor functions for individuals with disabilities to memory enhancement, direct brain-to-brain communication, and even thought typing.Its initial testing phase is directed towards individuals with paraplegia or tetraplegia, with the aim of empowering them to control smartphones or robotic limbs. Looking towards the future, Neuralink’s overarching goal is to achieve “symbiosis with artificial intelligence,” ensuring that humans can effectively keep pace with the ever-advancing field of AI. To achieve these ambitious objectives, Neuralink employs state-of-the-art robotics and flexible, bio-compatible materials to ensure high precision and throughput in the implantation of brain chips.

Advantages:

• Restoring Functionality for Disabilities: The ability to restore sensory and motor functions in individuals with disabilities, particularly those with paralysis, holds immense significance. Neuralink’s technology could significantly enhance their quality of life by providing greater control and autonomy.

• Memory Enhancement: The potential for memory enhancement is a ground breaking aspect of Neuralink’s technology. This could have profound implications for individuals with memory-related conditions or cognitive impairments, offering improvements in daily functioning and cognitive abilities.

• Symbiosis with AI: As technology continues to advance, achieving symbiosis with artificial intelligence becomes increasingly important. Neuralink’s goal to enable humans to integrate seamlessly with AI could be crucial for staying relevant in a world where AI capabilities are rapidly evolving.

However, Neuralink is not without its challenges. The company faces significant hurdles in developing safe and minimally invasive implantation procedures, obtaining regulatory approval for human trials, and navigating ethical concerns surrounding issues such as brain hacking, agency and identity, as well as data privacy. These considerations underscore the complex landscape in which Neuralink operates as it strives to bring its ground-breaking technology to fruition.

Disadvantages:

• Brain Hacking/Security Risks: One of the most critical disadvantages of Neuralink technology is the potential for brain hacking and security breaches. Implantable neural chips could be vulnerable to unauthorized access, allowing external parties to manipulate thoughts, memories, emotions, and movements. This poses significant security and privacy risks, raising concerns about the safeguarding of individuals’ most intimate cognitive processes.
• Unintended Side Effects: Implanting chips into the brain carries the risk of unintended side effects. The intrusion into the intricate neural landscape may unintentionally damage or disrupt healthy brain tissue and connections, leading to unforeseen neurological consequences. Addressing and mitigating these potential side effects poses a substantial challenge to the safety and viability of Neuralink’s technology.
• Dependency and Atrophy of Cognitive Abilities: Another major disadvantage revolves around the potential dependency on AI and external devices. Over-reliance on Neuralink’s technology may result in the underuse and atrophy of innate cognitive capabilities, as humans become increasingly dependent on the assistance provided by the implanted chips. This dependency could have profound implications for the overall cognitive health and self-sufficiency of individuals.

While the potential is exciting, these risks and concerns show the technology must be developed thoughtfully and deliberately, with a focus on ethics and safety. Oversight and public discussion should accompany progress. While still highly speculative, Neuralink demonstrates the potential for advanced brain-machine interfaces in the future. It remains to be seen if they can translate this vision into reality. The benefits could be profound but there are also risks around security, privacy and identity that need to be addressed responsibly. Neuralink represents a remarkable vision for the future of brain-machine interfaces. The possibilities of directly linking our biological brains with advanced AI systems are fascinating and could be truly transformative. However, this incredibly intimate merger of man and machine also raises major ethical, security, and societal concerns.

Conclusion:

In a nutshell, realised its full potential responsibly will require surmounting tremendous scientific and engineering challenges, comprehensive testing and regulatory approval for safety, and establishing policies that ensure the technology is used ethically. It remains to be seen whether Neuralink can achieve its lofty goals and avoid the pitfalls inherent in melding human brains with AI. While the possibilities are exciting, we must ensure human values, rights, and agency remain central to any technology advancements. If these challenges can be properly addressed, Neuralink could profoundly expand human potential and radically change our relationship with machines. But such a monumental transition requires prudence, care, and wisdom to put human interests first. With the right framework, diligent research, and open public discussion, we can work toward realizing the upsides of brain-machine interfaces while steering clear of the dangers. The future remains uncertain but also full of hope if we chart the course forward responsibly.

AI Tool: Claude

Prompts:

-What is neuralink?

-Advantages and Disadvantages

Reference:

https://www.usatoday.com/story/tech/2022/12/01/neuralink-elon-musk-brain-implant/10811520002/

https://builtin.com/hardware/what-is-neuralink

https://www.vox.com/future-perfect/23899981/elon-musk-ai-neuralink-brain-computer-interface

https://towardsdatascience.com/what-is-neuralink-a-look-at-what-it-is-what-it-wants-to-be-and-what-it-could-become-2acf32b51dc5

https://www.reuters.com/technology/musk-brain-implant-company-violated-us-hazardous-material-transport-rules-2024-01-26/

Reading Time: 3 minutes

Ever Wonder How Much Energy and Water are Consumed from AI Usage?

Well, a hidden but important worry in the fast-moving field of artificial intelligence (AI), where creativity has no boundaries, is energy consumption. Ever consider the environmental cost of AI’s unrelenting advancement? As AI develops, there is no denying the energy imprint it leaves behind, which begs ecological concerns. We explore how AI affects our resources in this blog.

Prediction of Future Energy Consumption

According to De Vries’ estimations, the projected manufacturing of AI servers could result in an annual increase in global electricity consumption connected to AI of 85 to 134 TWh by 2027. This would be like what nations like the Netherlands, Argentina, and Sweden would require each year in electricity. The increasing energy consumption of artificial intelligence (AI) raises important questions regarding sustainability and forces a closer look at the environmental effects of this rapidly developing industry.

Why Does AI Use so Much Energy? 

The enormous volume of data that AI models are trained on, their complexity, and the sheer number of queries that users make of them are the main causes of their excessive energy consumption. Training an AI model can take a few minutes to several months, depending on the model’s complexity and volume of data, according to Lyline Lim, Head of Impact and Sustainability at PhotoRoom, a top AI photo editor.

Factors of Energy Consumption

• Data-Intensive Training: Data training takes up extensive amount of computation power. This drives up energy consumption significantly.
• Algorithmic Complexity: The more complex and advanced the algorithms is, the higher the energy consumption. 
• Hardware Efficiency: The rate of complexity of AI models outpaces the improvement in hardware efficiency. Ultimately, the overall energy usage still increases despite a more efficient hardware.

Examples of Some Industries Influencing Consumption

• Autonomous systems: (E.g. robotics/ drones) Rely significantly on artificial intelligence (AI) to make decisions, which requires a lot of processing power.
• Medical Application: AI is using more energy as its use in medical imaging and diagnostic technologies expands.
• Cybersecurity: AI-powered fraud detection and risk analysis solutions in a variety of industries greatly increase energy usage.
• Marketing: Popularity of recommendation systems and personalized advertising driven by AI

Water Usage

AI processing servers pose a twin environmental problem because to their high energy requirements and large water usage for cooling systems. The need for AI globally might account for up to 6.6 billion cubic meters of water consumption in 2027. That is more than half of the United Kingdom’s annual water withdrawal total.

What are some mitigating solutions?

Creating more energy-efficient hardware, streamlining data, putting money into green data centers, embracing responsible AI use, and creating precise guidelines for tracking and disclosing AI’s energy footprint to lessen the technology’s negative environmental effects. This will stimulate innovation in energy-efficient AI systems and help to advance accountability.

Conclusion

This blog discusses the effects of using AI while also pointing out the major advantages it has for customers. Even while they might not have direct control over change, customers still have a voice and can put pressure on businesses. Customers can play a crucial role in guiding the industry towards a more ethical path by pushing for greater transparency and pressuring businesses to invest in more sustainable AI methods that take the responsible use of natural resources into consideration.

AI: You.com

Prompt: Energy and water consumption from AI usage

https://www.bbc.com/news/technology-67053139

https://www.euronews.com/next/2023/10/10/demand-for-ai-could-mean-technology-consumes-same-energy-as-a-country-analysis-shows

https://www.forbes.com/sites/sap/2023/12/20/why-does-ai-consume-so-much-energy/#

https://www.scientificamerican.com/article/the-ai-boom-could-use-a-shocking-amount-of-electricity/

https://oecd.ai/en/wonk/how-much-water-does-ai-consume

Reading Time: 3 minutes

The term “metaverse” refers to a collective virtual shared space that is created by the convergence of physical and virtual reality. It’s a space where users can interact with a computer-generated environment and other users in real-time. The metaverse was once the next big thing in technology, with Facebook even changing its name to Meta Platforms to reflect its commitment to this virtual world. Metaverse was popular in the eCommerce arena, with Gucci, Prada and Louis Vuitton being the first mover.

To celebrate the 200th birthday of Louis Vuitton’s founder, the luxury brand unveiled a mobile game. In this interactive experience, players can delve into six distinct Metaverse realms. Discovering the rich history of LV. Participants have the opportunity to acquire virtual branded collectibles and exclusive in-game NFTs, offering a novel means for the brand to pay homage to its profound heritage and enduring legacy. This innovative approach not only enhances the customer experience but also forges a stronger connection between the audience and the brand.

But, what happened to the metaverse? Metaverse started losing its hype following the crash of cryptocurrency as funds for the project depleted. Nevertheless, there are additional factors contributing to the potential decline of the metaverse.

• User Adoption

Difficulties including hardware requirements, intricate user interfaces, and a dearth of interesting applications have kept the typical user from being very interested in or involved in the Metaverse. The growth of the Metaverse has been impeded by the lack of a critical mass of active users, which has prevented the establishment of dynamic virtual communities and businesses.

• Digital Inequality

There have been major obstacles in the form of network capabilities, processing power, and bandwidth constraints due to the infrastructure needed to handle such a large scale. It is reported that users experience a fragmented experience as there were constant lags on the frame rates and overall a buggy experience, disrupting the seamless experience.  The full potential of the Metaverse has not been realised because of these technical limitations. 

• Ethical Concerns

The metaverse gathers a lot of data about identity, data privacy, social and economic inequality, safety and content, addiction and mental health, child protection, and environmental sustainability. It can be difficult to determine who owns this data the developer or the brand. Significant gaps are created by this, which could harm end users. Further improvements to existing laws are necessary to ensure the metaverse develops in a responsible and equitable manner.

In conclusion, after cryptocurrency crashes and followed by financing problems, the metaverse which had previously seemed like a promising technological frontier saw a decrease in appeal. Significant obstacles to its further growth include issues like user acceptance difficulties, digital inequality leading to technical limits, and ethical concerns over data ownership. Although companies such as Louis Vuitton have embraced the metaverse in a creative way, these problems nonetheless limit the metaverse’s total potential. It is imperative that the metaverse tackles these issues with enhanced technology, infrastructure, and legislation in order to realise its potential for responsible and fair development.

AI: Perplexity (Prompt: What happened to the Metaverse?)

https://dmexco.com/stories/welcome-to-the-metaverse-of-e-commerce/#:~:text=%232%20The%20metaverse%20as%20an%20arena%20for%20virtual%20brand%20building&text=Gucci%2C%20Prada%2C%20Clinique%2C%20and,to%20give%20users%20memorable%20experiences.

https://www.skrill.com/en/skrill-news/crypto/what-happened-to-the-metaverse/#:~:text=Unfortunately%2C%20when%20the%20crypto%20market,focused%20projects%20crashed%20with%20it.

https://www.forbes.com/sites/jackkelly/2022/02/03/cant-be-evil-the-democratized-vision-of-the-metaverse-enabled-decentraland-to-flourish/

https://cointelegraph.com/news/the-ethics-of-the-metaverse-privacy-ownership-and-control

https://www.fisclouds.com/decoding-the-downfall-of-metaverse-and-nft-contrasting-paths-for-the-ai-trend-9708/#:~:text=Factors%20contributing%20to%20the%20downfall%20of%20the%20Metaverse&text=The%20infrastructure%20required%20to%20support,of%20the%20Metaverse%27s%20full%20potential.

Reading Time: 3 minutes

Intrigued by the ever-evolving landscape of AI, I recently came across Bard by Google from one of my module. This platform had eluded my awareness until then. So I thought it might be interesting to explore Bard in this blog post. We have came across countless AI operated chat platforms and in 10 days it will be chatGPT’s 1 year old birthday! ChatGPT was a game changer for everyone, for better or worse. After the launch of chatGPT, many others AI chatbots appeared for and is free for public use.

What is Bard?

Bard is a large language model developed by Google AI. Based initially on Language Model for Dialougue Applications (LaMDA) a family of large language models (LLMs) and later moved on to PaLM 2. A next generation of language model that is able to breakdown complex task, and better in human nuances compared to previous model. It functions similarly to chatGPT, but even better because it connects to the internet unlike chatGPT 3.5, which plugins are required. 

Bart vs Chatgpt

• Better UX
• Enables uses to edit their responses without needing to enter a new prompt.
• Insert preference in conversations (e.g. shorter/ longer text, formal/ casual) whereas chatGPT you have to prompt. 

• Wallet Friendly
• Free access to the internet unlike chatGPT which limits this function to their paid users.

• Build for Research Purposes
• Click on “Double-check response” (Green Highlight = Similar Information & Orange Highlight= Differing information)
• Offers related search queries.
• Provides 3 drafts for user to choose the best ones.

Cons of Bard

• Reliability 

Sources provided by Bard may not be reliable, so users must always cross check the references and evaluate whether it’s dependable.

• Stuck in Googles Ecosystem

It an isolated experience where information comes from only google. Whereas, for chatGPT you are able to integrate with external platforms (Plugins).

• Prone to “Hallucination”

This is a common problem with both ChatGPT and Bard. This is when inaccurate answers are generated. There are several causes such as low quality data, insufficient data and prompts that are designed to mislead it.

In conclusion, AI is here to make our life’s more efficient prompt us when we hit a mental block. We must always be cautious to fact check our information.

In conclusion, AI enhances efficiency and aids us during mental blocks. While it improves our lives, it’s crucial to exercise caution, verifying information to mitigate potential risks. As we embrace AI’s benefits, maintaining a vigilant approach ensures responsible and informed utilization of these technologies in our daily lives.

References:

https://www.vox.com/future-perfect/23617185/ai-chatbots-eliza-chatgpt-bing-sydney-artificial-intelligence-history

https://ai.google/discover/palm2/#:~:text=PaLM%202%20is%20grounded%20in,models%2C%20like%20Sec%2DPaLM.

https://zapier.com/blog/chatgpt-vs-bard/

https://www.vox.com/future-perfect/23617185/ai-chatbots-eliza-chatgpt-bing-sydney-artificial-intelligence-history

https://zapier.com/blog/ai-hallucinations/

AI Platform: Google Bard

– What is Google Bard?

– What is Bard’s competitive edge compared to chatGPT?

Reading Time: 3 minutes

Introduction

According to FICO, in Poland, fraud levels have increased by over 300%, posted as the fourth-highest loss percentage increase in Europe in 2021. In the ever-evolving landscape of technology, the digital realm has revolutionized the way we live, work, and connect. These digital platforms have brought us great convenience and opportunities, whether it is connecting with your loved ones or dealing with everyday responsibilities. While technology has enriched our lives in countless ways, it simultaneously offers fertile ground for individuals with malicious intent. These criminals have gained an upper hand in deceiving anyone into their trickery through their sophisticated tactics.

Types of Technology-Enabled Scams:

•          Online phishing attacks

•          Email and social engineering scams

•          Online shopping fraud

•          Identity theft

•          Cryptocurrency scams

What enables fraud to be carried out efficiently:

Social Engineering: Scammers would employ multiple persuasive and cognitive bias techniques. They can make you feel overwhelmed and unable to process your thoughts coherently and with a little bit of pushing you would just fall into their demands. This emotional factor is the foundation of falling prey to a scam.

Digital Identity App

In Singapore, they have an identity app called “Singpass” which could substitute our ID for official purposes. It provided its residents with great convenience as we were able to open bank accounts and settle governmental-level administrative stuff through this app. This had certainly improved efficiency on administrative matters but at what cost?

Scammers were able to convince their victims to hand over their accounts through emotional manipulation. With access to this app, they were able to access banking functions that enabled their fraudulent activities to be untraceable. They will launder money to one victim’s account and another and the cycle goes on. In case you are wondering, there are security measures in place but sometimes victims themselves willingly give their One-Time passwords away usually after some emotional manipulation.

Links: https://www.channelnewsasia.com/singapore/singpass-account-8000-launder-220000-scam-victims-court-3825921

E-wallet:

A digital wallet facilitates seamless financial transactions via a mobile device. The integration of digital wallets has been transformative. Personally, I love it! I do not have to deal with exorbitant bank transfer rates or unideal exchange rates at the money changer when travelling to another country. However,  it is a double-edged sword as now users are prone to phishing sites, malware that could steal your payment information Wi-Fi-snooping etc. This is one of the platforms where victims could set up an account easily and money to be transferred to the fraudulent party instantly. It is impossible to cancel it, leaving victims at a loss.

Links: https://news.yahoo.com/posts-target-indonesia-digital-payment-091427637.html

Which party should be responsible?

No one wishes to fall prey to scams, but what happens if you do? Recently, the Singapore government entities have proposed that financial institutions and telecommunication providers should share the responsibility for phishing scam losses. Although, it is still in its initial stages. I think it is a good strategy for organisations to take responsibility to a certain extent.

Link: https://www.channelnewsasia.com/singapore/financial-institutions-telcos-share-liability-consumers-scam-losses-mas-framework-3871356

Conclusion

These risks are inevitable when living in a world that relies on the Internet of Things so readily. Individuals, governments and organisations should be wary and alert of the consequences of enjoying such convenience. As technology advances, so do cybercriminal tactics. Safety in the digital world is a shared responsibility. Balancing convenience with security requires constant vigilance and collaboration among individuals, organizations, and governments

ChatGPT 3.5:  “What kind of sub-topics can I write for How technology empowers fraudulent schemes?”

Links: https://www.linkedin.com/pulse/why-even-young-savvy-well-educated-fall-scams-stephen-le-hoang-chuong#:~:text=Scammers%20are%20successful%20because%20they,an%20effective%20and%20common%20one.

A Guide for Foreigners to setting up a Singpass account

Digital Wallet Fraud – What is it and how it’s prevented

Financial Institutions to Bear Full Losses in Phishing Scams: A Paradigm Shift in Accountability

https://www.fico.com/europeanfraud/poland