Category Archives: Artificial General Intelligence

Senior Minister of State for Defence Mr Heng Chee How officiated the inaugural AI Student Developer Conference at the Lifelong Learning Institute today. Organised by AI Singapore (AISG) and attended by more than 300 participants, the conference allowed attendees to gain insights into Artificial Intelligence (AI) and the AI industry through panel discussions, interactive booths and workshops, as well as explore career opportunities with industry partners. As part of the conference, a Memorandum of Understanding (MOU) between AISG and the Singapore Armed Forces (SAF)'s Digital and Intelligence Service (DIS) was signed.

Delivering the opening address at the conference, Mr Heng said that, "Outside of the defence-specific sector partners, DIS is also enlarging its engagement with the wider technology ecosystem, including engagement with the commercial sector and academia This MOU is another example of DIS's pursuit in this direction of engagement, augmenting our ongoing efforts to build and sustain a strong and capable workforce and talent pipeline to strengthen and sharpen the SAF's digital cutting edge."

The MOU between AISG and the DIS was signed by Head of LearnAI at AISG, Mr Koo Sengmeng and DIS Chief Digitalisation Officer Military Expert 7 (ME7) Guo Jinghua. Senior Director of AI Governance at AISG, Prof Simon Chesterman and Chief of Digital and Intelligence Service/Director Military Intelligence, Brigadier-General Lee Yi-Jin witnessed the signing of the MOU, which formalises the collaboration in deepening national AI expertise for Singapore's digital defence.

The MOU will further collaboration and strengthen the DIS's capability development in Data Science and AI (DSAI). The DIS will need to keep pace with, and agilely harness the rapid pace of AI innovation in academia and industry, to complement the strong AI capabilities of the Defence Technology Community. This is crucial for the DIS to better exploit the vast and growing volume of data in the digital domain, and effectively detect and respond to the increasing digital threats facing Singapore and Singaporeans. The DIS will leverage AISG's industry and talent development programmes including the 100 Experiments (100E) and AI Apprenticeship Programme (AIAP) to expand the DIS's capacity to deploy advance AI techniques, such as the use of Large Language Models and Reinforcement Learning, and integrate them into operations of the DIS and the SAF.

The DIS will also work with AISG to develop and expand its workforce. Through the introduction of AISG's LearnAI courses, the DIS will expand its course offerings for DIS personnel's professional upskilling. The DIS will also leverage AISG's existing networks of students to sustain the DSAI talent pipeline, while supporting AISG's mandate of growing and developing a national digital workforce. The DIS will enable national talents in AISG's AIAP, who are undergoing AI deep-skilling, to contribute to national defence via their involvement in the various projects supporting the DIS. The DIS will also offer employment opportunities to these talents where suitable. In addition, AISG will share about National Service (NS) and career opportunities in the DIS, such as the Digital Work-Learn Scheme[1], with students from the AISG Student Outreach Programme.

Highlighting the importance of the MOU for Singapore's digital defence, Mr Koo said, "Our partnership with the DIS will ensure that Singapore has a robust and resilient pipeline of AI talents that have knowledge of issues related to national defence and possess the relevant expertise to protect our digital borders and safeguard Singapore. We look forward to working closely with the DIS to collectively deepen the core competencies of our next-generation Singapore Armed Forces to stay ahead of the threats of tomorrow."

ME7 Guo said, "The DIS and AISG are working towards our common goal of strengthening digital capabilities to safeguard Singapore. The effective use of AI is crucial for the SAF's mission success. We need to better reap the dynamic AI innovations in academia and industry, and integrate them into SAF operations. Our partnership with AISG is therefore an important part of our approach to leverage cutting-edge AI innovations. Beyond AI capability development, our partnership with AISG will help grow the DIS digital fighting force to defend Singapore in the digital domain, and contribute to the national AI talent pipeline through various schemes as the Digital Work-Learn Scheme."

[1]Servicemen under the WLS will serve for four years as Digital Specialists in the SAF, in a combination of full-time National Service and Regular service, developing data science, software development and AI skills through vocational, on-the-job and academic training.

About AI Singapore

AI Singapore (AISG) is a national AI programme launched by the National Research Foundation (NRF), Singapore to anchor deep national capabilities in artificial intelligence (AI) to create social and economic impacts through AI, grow the local talent, build an AI ecosystem, and put Singapore on the world map.

AISG brings together Singapore-based research institutions and the vibrant ecosystem of AI start-ups and companies developing AI products to perform applications-inspired research, grow the knowledge, create the tools, and develop the talent to power Singapore's AI efforts.

AISG is driven by a government-wide partnership comprising NRF, the Smart Nation and Digital Government Office (SNDGO), Economic Development Board (EDB), Infocomm Media Development Authority (IMDA), SGInnovate, and the Integrated Health Information Systems (IHiS).

Details of some of its programmes can be found below:

-100 Experiments (100E)

-AI Apprenticeship Programme (AIAP)

-LearnAI

For more information on AISG and its programmes, please visit: http://www.aisingapore.org

AI Singapore's Social Media Channels:

Facebook: https://www.facebook.com/groups/aisingapore

Instagram: @ai_singapore

LinkedIn: https://www.linkedin.com/company/aisingapore/

Twitter: https://twitter.com/AISingapore

About The DIS

As part of the transformation of the Next Generation SAF, the Digital and Intelligence Service, the fourth Service of the Singapore Armed Forces (SAF) was established in 2022. The DIS sees the consolidation and integration of existing Command, Control, Communications, Computers and Intelligence (C4I) as well as cyber capabilities of the SAF. As a dedicated Service, the DIS will raise, train and sustain digital forces and capabilities to fulfil its mission to defend the peace and security of Singapore from the evolving and increasingly complex threats in the digital domain.

The mission of the DIS is to defend and dominate in the digital domain. As part of an integrated SAF, the DIS will enhance Singapore's security, from peace to war. The DIS plays a critical role in defending Singapore from threats in the digital domain, and allows the SAF to operate better as a networked and integrated force to deal with a wider spectrum of external threats to enhance and safeguard Singapore's peace and sovereignty. The DIS collaborates with partners across the MINDEF, SAF, Whole-of-Government agencies and like-minded partners in academia and industry in defending our nation against threats in the digital domain.

Building a highly-skilled digital workforce is key to the digital defence strategy of the SAF. The DIS continually attracts and develops both military and non-uniformed digital experts to grow the SAF's digital workforce.

The DIS leverages our National Servicemen to develop its digital workforce. Operationally Ready National Servicemen (ORNS) with matching talents and relevant civilian expertise may also express interest to serve in the DIS through the Enhanced Expert Deployment Scheme (EEDS). Full-time National Servicemen (NSFs) with suitable skills are offered to participate in DIS-related Work-Learn Schemes (WLS) where they will be able to undergo military training and serve NS while attaining academic credits which will contribute to the eventual completion of a relevant university degree. There are currently two DIS WLS, namely the Digital WLS and Cyber WLS.

For more information on the DIS and its careers, please visit: http://www.mindef.gov.sg/dis

The Digital and Intelligence Service's Social Media Channels:

Facebook: https://www.facebook.com/thesingaporeDIS

Instagram: @thesingaporedis

LinkedIn: https://www.linkedin.com/company/digital-and-intelligence-service

Twitter: @thesingaporeDIS

Go here to read the rest:

AI Singapore and the Digital and Intelligence Service Sign ... - MINDEF Singapore

ARTIFICIAL intelligence has been all over the news in the past few years. Even so, in recent months the drumbeat has reached a crescendo, largely because an AI-powered chatbot called ChatGPT has taken the world by storm with its ability to generate fluent text and confidently answer all manner of questions. All of which has people wondering whether AIs have reached a turning point.

The current system behind ChatGPT is a large language model called GPT-3.5, which consists of an artificial neural network, a series of interlinked processing units that allow for programs that can learn. Nothing unusual there. What surprised many, however, is the extent of the abilities of the latest version, GPT-4. In March, Microsoft researchers, who were given access to the system by OpenAI, which makes it, argued that by showing prowess on tasks beyond those it was trained on, as well as producing convincing language, GPT-4 displays sparks of artificial general intelligence. That is a long-held goal for AI research, often thought of as the ability to do anything that humans can do. Many experts pushed back, arguing that it is a long way from human-like intelligence.

So just how intelligent are these AIs, and what does their rise mean for us? Few are better placed to answer that than Melanie Mitchell, a professor at the Santa Fe Institute in New Mexico and author of the book Artificial Intelligence: A guide for thinking humans. Mitchell spoke to New Scientist about the wave of attention AI is getting, the challenges in evaluating how smart GPT-4 really is, and why AI is constantly forcing us

Here is the original post:

How smart is ChatGPT really and how do we judge intelligence in AIs? - New Scientist

ChatGPT seems to be everywhere. From in-depth reports in highly respected technology publications to gushing reviews in mainstream media, ChatGPT has been hailed as the next big thing in artificial intelligence, and with good reason.

As a developer resource, ChatGPT is simply outstanding, particularly when compared to searching existing resources such as Stack Overflow (which are undoubtedly included in GPTs data model). Ask ChatGPT a software question and you get a summary of available web solutions and some sample code that can be displayed in the language you need. Not happy with the result? Get a refined answer with just a little added info as the system remembers the context of your previous queries. While the just-released GPT-4 offers some significant new features, its usefulness to a developer hasnt changed much in my usage.

As a software asset, ChatGPTs API can be used to give the illusion of intelligence to almost any interactive system. As opposed to typing questions into the web interface, ChatGPT also offers a free API key which enables a program to ask questions and process answers. The API also provides access to features that are not accessible via the web, including options like how long an answer is expected and how creative it should be.

But while ChatGPT has already attracted more than a hundred million users, drawn by its impressive capabilities, it is important to recognize that it only gives the illusion of understanding. In reality, ChatGPT is manipulating symbols and code samples which it has scoured from the web without any understanding of what those symbols and samples mean. If given clear, easy questions, ChatGPT will offer (usually) clear, accurate responses. If asked tricky questions or questions with false or negative premises, the results are far less predictable. ChatGPT can also provide plausible sounding, but incorrect answers and can often be excessively verbose.

So whats wrong with that? To a developer, not much. Simply cut-and-paste the sample code, compile it, and youll know in a few seconds whether or not the answer works properly. This is a different situation than asking a health question, for example, where ChatGPT can report data from dubious sources without citing them, and it is time-consuming to double-check the results.

Further, the new GPT-4 system isnt very good a working backwards from a desired solution to the steps needed to achieve it. In a programming context, we are often given an existing data set and a desired outcome and need to define the algorithm to get from one to the other. If such an algorithm already exists in GPTs dataset, it will likely be found and modified to fit the needed capabilities. Great for a majority of instances. If a new algorithm is needed, though, GPT should not be expected to define one.

ChatGPT represents an incredibly powerful tool and a major advance in self-learning AI. It represents a step toward artificial general intelligence (AGI), the hypothetical (though many would argue inevitable) ability of anintelligent agentto understand or learn any intellectual task thata human can. But it makes only a pretense of actual understanding. It simply manipulates words and symbols. In fact, AI systems such as ChatGPT may be slowing the emergence of AGI due to their continued reliance on bigger and more sophisticated datasets and machine learning techniques to predict the next word or phrase in a sequence.

To make the leap from AI to AGI, researchers ultimately must shift their focus to a more biologically plausible system modeled on the human brain, with algorithms that enable it to build abstract things with limitless connections and context, rather than the vast arrays, training sets, and computer power todays AI demands.

For AGI to emerge, it must have the capability to understand that physical objects exist in a physical world and words can be used to represent those objects, as well as various thoughts and concepts. Because concepts such as art and music, and even some physical objects (those for example which have tastes, smells, or textures) dont easily lend themselves to being expressed in words, however, AGI must also contain multisensory inputs and an underlying data structure which will support the creation of relationships between multiple types of data.

Further, an internal mental model of the AGIs environment with the AGI at its center is essential. Such a model will enable an artificial entity to have perspective and a point of view with respect to its surroundings that approximates the way in which humans see and interpret the world around them. After all, how could a system have a point of view if it never experienced one?

The AGI must also be able to perceive the passage of time, which will allow it to comprehend how each action it takes now will impact the outcomes it experiences in the future. This goes hand-in-hand with the ability to exhibit imagination. Without the ability to imagine, AGI will be incapable of considering the numerous potential actions it can take, evaluating the impact of each action, and ultimately choosing the option that appears to be most reasonable.

There are certainly other capabilities needed for AGI to emerge, but implementation of just these concepts will allow us to better understand what remains to be done for AGI to be realized. Moreover, none of these concepts are impossible to create. To get there, though, researchers need to abandon the current, widely used model of extending a text-based system like ChatGPT to handle multisensory information, a mental model, cause-and-effect, and the passage of time. Instead, they should start with a data structure and a set of algorithms and then utilize the vision, planning, and decision-making capabilities of an autonomous robot to extend these capabilities to ChatGPTs text abilities.

Fortunately, a model for doing all these things already exists in an organ which weighs about 3.3 pounds and uses about 12 watts of energythe human brain. While we know a lot about the brains structure, we dont know what fraction of our DNA defines the brain or even how much DNA defines the structure of its neocortex, the part of the brain we use to think. If we presume that general intelligence is a direct outgrowth of the structure defined by our DNA and that structure could be defined by as little as one percent of that DNA, though, it is clear that the real problem in AGI emergence is not one that requires gigabytes to define, but really one of what to write as the fundamental AGI algorithms.

With that in mind, imagine what could happen if all of todays AI systems were to be built on a common underlying data structure which would enable them and their algorithms to begin interacting with each other.Gradually, a broader context that can understand and learn would emerge. As these systems become more advanced, they would slowly begin to work together to create a more general intelligence that approaches the threshold for human-level intelligence, then equals it, then surpasses it. Perhaps only then will we humans begin to acknowledge that AGI has emerged. To get there, we simply need to change our approach.

Portions of this article are drawn from Microsoft Researchs just-published paper, Sparks of Artificial General Intelligence: Early experiments with GPT-4 By Sebastien Bubeck, et al, https://arxiv.org/pdf/2303.12712.pdf

Excerpt from:

ChatGPT is impressive, but it may slow the emergence of AGI - TechTalks

The partnership between SingularityNET and Simuli.ai aims to speed up artificial general intelligence advancement, and will focus on the creation of a Metagraph Pattern Matching Chip (MPMC).

This new chip will host the two knowledge graph search algorithms, Breadth-First Search (BFS) and Depth-First Search (DFS).

Combining these building blocks for AI systems into one chip can enable more intuitive knowledge representation, reasoning and decision-making.

Once created, the MPMC will be integrated with Simulis pre-existing Hypervector Chip which is used for processing data patterns with fewer processors than traditional hardware to create an AGI board that aims to accelerate realisation of artificial general intelligence capabilities.

The new hardware is set to be utilised by SingularityNETs spin-off project TrueAGI to offer AGI-as-a-service to enterprise organisations.

Together, SingularityNET and Simuli.ai aim to mitigate common hardware constraints faced by AI developers, such as being limited to graphics processing units (GPU), which help devices handle graphics, effects and videos.

In addition, the project looks to lower the cost of AI training and interference by allowing for this to be achieved with less required hardware.

The Simuli AGI board has strong potential to catalyse emergence of a new era of AI techniques and functions, said Dr. Ben Goertzel, CEO of SingularityNET.

The core of what we need to progress from narrow AI to AGI is of course the right cognitive architectures and learning and reasoning algorithms but without the right hardware, even the best mathematics and software cant run efficiently enough to have practical impact.

So many AI methods weve been working on for decades, are going to finally be able to show their stuff in a practical sense when given the right hardware to run on.

Rachel St.Clair, CEO of Simuli.ai, commented: The power of optimising large scale AGI models to run faster by leveraging Simulis hardware platform is multifold.

First, these AGI frameworks get rapid development that wasnt exactly possible without large compute cost prior. Then, such a device as the AGI motherboard can expand the types of code that can be run scalably and efficiently in a single instance of the AGI model, for example Hyperon.

Also, scalable computing is better for longevity of the planet and the technology itself, so optimising on both the SW/HW sides is key. This will likely result in AGI thats better for everyone. Were excited to be playing a role in tipping the scale from AI to AGI.

Artificial general intelligence a forward-looking term referring to machine intelligence that can solve problems and complete tasks to the same standard as a human has been cited as a next step in AI development, with generative AI being the most prominent current innovation trend in the space. Read more about AGI here.

Why CIOs are turning to knowledge graphs for critical business help Exploring the value of knowledge graphs for chief information officers (CIOs).

What ChatGPT means for developers Will ChatGPT replace software developers, or will it make their jobs more enjoyable by doing the heavy lifting when it comes to writing boilerplate code, debugging and testing?

Visit link:

New AGI hardware in progress for artificial general intelligence - Information Age

Towards artificial general intelligence, ChatGPT 5 is on track and has already begun to trend

Towards artificial intelligence and general intelligence, ChatGPT 5 is on track and has already begun to trend on Twitter, with many people guessing about the ChatGPT platforms future evolution. These ChatGPT-5 Twitter debates have gotten even more interest than those over ChatGPT-4. Many users have great hopes for the future edition, believing that it would feature immaculate visuals, something that ChatGPT-4 has yet to achieve.

OpenAI is actively developing new features for ChatGPT and intends to release GPT-5 later this winter. According to studies on GPT-5s capabilities, OpenAI may be on the verge of reaching Artificial General Intelligence (AGI) and becoming practically indistinguishable from a person in its capacity to create natural language answers.

While ChatGPT may become indistinguishable from a person in terms of natural language answers, it will still outperform the human brain in terms of data processing and content creation. ChatGPT already gained considerable new features as a result of the recent upgrade to GPT-4, enhancing the chatbots utility as a tool. ChatGPT now supports multimodal input, allowing it to receive data via text and graphics and produce replies in many languages.

Some intriguing tweets indicate that ChatGPT-4 has spontaneously addressed ChatGPT-5, even when not pushed to do so. This begs the issue of what ChatGPT-4 is aware of regarding the next version that we are not. Furthermore, GPT-4 has shown exam-taking abilities that outperform those of its predecessor. One of the developers, Siqi Chen, stated on Twitter that GPT-5 will complete its training by December, with OpenAI expecting it to attain AGI.

Whether or not GPT-5 achieves AGI, it should offer major enhancements above GPT-4, which was already a huge advance for ChatGPT. Its impossible to foresee the entire scope of these enhancements, but the chatbot might allow different input modalities and produce faster and more accurate answers. While the possibility of ChatGPT improving and achieving AGI is intriguing, we must equally examine the potential negative implications.

This revelation is expected to ignite a heated argument about whether GPT-5 has genuinely reached AGI, and given the nature of such disputes, it is quite likely that it will be considered to have gained AGI. It implies that with the aid of GPT-5, Generative AI might achieve human-like indistinguishability. Chen further emphasized on Twitter that, while reaching AGI with GPT-5 is not a majority opinion within OpenAI, some people feel it is doable. If Artificial Intelligence achieves AGI, it will have intellectual and task comprehension abilities equivalent to humans.

It is impossible to forecast what these negative consequences would be, just as it is difficult to envision the good consequences of ChatGPT attaining AGI. Despite this ambiguity, there is no need to be concerned about a sci-fi movie scenario in which AI takes over. However, the growth of AI has already prompted worries at Europol, since criminals are exploiting the capabilities of non-AGI versions of ChatGPT for illicit purposes. Before the introduction of GPT-5, we may see an interim version of ChatGPT.

Read more here:

Towards Artificial General Intelligence, ChatGPT 5 is on Track - Analytics Insight

By Veronica Brinkley

As AI models have advanced, it has become increasingly evident that they will play an important role in the future of humankind. Current models have a range of capabilities that are supposedly designed to aid humans. For instance, photo-generating models like MidJourney and DALL-E have the ability to create images in a multitude of styles, based on the users prompt. These programs outputs have become increasingly accurate to the point that, to the average viewer, they are often indiscernible from real photographs.

The language model ChatGPT is garnering the most media attention. ChatGPT is advancing rapidly; its already on its fourth version. The lab behind it, OpenAI, has stated that the company is building towards an ambitious goal artificial general intelligence (AGI), their term for an AI that is as smart as, if not smarter than, the average human.

These developments have raised alarms in the tech community. Recently, over 1,000 major tech executives including Elon Musk professors, and scientists signed an open letter directed toward OpenAI, requesting an immediate sixmonth pause in artificial intelligence development. The main concern posited by the letter is that AI systems with human-competitive intelligence can pose profound risks to society and humanity, and therefore necessitate governmental regulation. The letter went on to say that AI development labs are locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one not even their creators can understand, predict, or reliably control.

OpenAI, for its part, says its technology will improve society. According to their website, advancements in AI could help us elevate humanity by increasing abundance, turbocharging the global economy, and aiding in the discovery of new scientific knowledge that changes the limits of possibility. To me, this just sounds like a lot of buzzwords and it doesnt really say much about what they intend to do with their product.

Now, if youre like me, youre probably thinking, I swear Ive seen this in a movie, and it did not end well. Its scary to see the beginning of the march to machine intelligence. Science fiction centered around AI previously felt abstract, but now seems potentially accurate. My personal favorite example is the dystopian sci-fi video game Detroit: Become Human.

Quantic Dreams Detroit: Become Human is set in the not-so-distant future of 2038, in an America where highly-developed androids have become commonplace. The economy is completely dependent on them as the means of production. However, the androids begin to gain sentience and deviate from their programming. This sparks a civil rights movement of deviant androids. A power struggle ensues, as humans refuse to accept androids as autonomous beings. While current AI is far from this reality, it is a chilling projection of what could be in store. The game itself directly comments on this reality in its opening lines: Remember, this isnt just a game, its our future.

Drawing parallels between the story of Detroit: Become Human and our current social trajectory is hardly difficult. CyberLife, the AI research and development firm in the games setting, represents a potential future for OpenAI. In the game, CyberLife has become the standard for androids and therefore holds an immensely disproportionate amount of power over the economy and ruling bodies. Perhaps we arent as far away from this reality as we think. In order to prevent such a future, industries need to change to accommodate AI, a technology that is only growing faster, smarter and more powerful. This is where the government must step in. It must regulate the creation of AI very consciously.

In a perfect world, state leaders consciously and unerringly regulate the creation of AI, acting free from considerations of profit and power. However, as we know, the government doesnt have a great history of neutrality or altruism. And few in Washington even understand technology just watch the Congressional hearings on Facebook or TikTok for examples. Washington has already failed to stay ahead of tech decisions that affect millions. OpenAI has been seemingly honest about these concerns, stating, we hope for a global conversation about three key questions: how to govern these systems, how to fairly distribute the benefits they generate, and how to fairly share access. While its great that they hope for this outcome, I do not believe that hoping is enough. Improper handling and regulation could have catastrophic effects on society, as seen in the manipulation and misuse of current social media. Detroit: Become Human may not be far off.

As college students who are just beginning to enter the job market, these advancements could easily affect us in the near future. Entry-level positions might be displaced by AI and it may become increasingly difficult for us to find meaningful work. In terms of possible issues with the use of this technology, this is just the tip of the iceberg.

So many of us grew up watching events like this happen in the movies and on TV, and it is hard to believe what was once science fiction is beginning to exist. Its also alarming to watch it unfold, knowing it could impact our futures. But we are not helpless. We can stay informed about technological advancements. We can slow their deployment until the ramifications are understood. We can apply the lessons learned from fictional media and the real-life corruption of social media. And we must consider enacting accompanying regulations on tech industries. We should be mindful that technology is not always a wonder especially in the hands of mere mortals.

Visit link:

Is artificial intelligence approaching science fiction? - The Trail - The Puget Sound Trail