The transformative potential of artificial intelligence (AI) is reshaping our world, presenting both thrilling opportunities and profound challenges. In the Philippines, where Dominic Ligot (Founder, AI expert & communicator)[1]and I have delved into AI’s possibilities, the integration of this technology into our unique cultural fabric can unleash exponential productivity. This post explores how we (as a country) can harness AI to enhance our local community without losing our distinct identity. With our shared fascination as a spark, I’ll examine the broader implications of AI, guided by the imperative to understand its complexities and the need to proceed with caution and adherence to alignment research[2].
Transformative Technologies and AI Research
The Transformative Potential of AI
AI stands as a transformative technology, holding the potential to reshape our world in the coming decade. The failure to understand and integrate AI could leave a country behind in a rapidly changing landscape.
Historical Precedence of Technological Transformations
Technological advancements have driven significant transformations. Whether through the application of cutting-edge physics in warfare or the discovery of new scientific fields, technological innovations have often led to pivotal moments in human history.
The Complexity and Urgency of AI Research
AI research is neither simple nor conceptually straightforward. The very complexity of the field underscores the importance of pursuing a deep understanding of it. While the path towards mastering AI may seem daunting and even wild, the stakes are too high for complacency.
The Inevitable Direction of AI Research
Our current situation leaves us with few choices but to embrace AI research. It’s not merely an option; it’s an essential direction we must take to ensure our survival in a future increasingly shaped by artificial intelligence. The question is not whether we should pursue AI research, but how we can do so most effectively and responsibly.
Having explored the immense transformative power of AI and its historical parallels, it’s essential to consider both the challenges and promises this complex field presents to us today.
Upsides and Downsides
AI Research Challenges and Uncertainties
AI research at present resembles alchemy more than mathematics. We still lack a formal theory on the malleability of data that can be infused into models, making our understanding of how data gets distributed to individual parts, such as tokens or layers, uncertain. These distributional shifts affect how technology can respond in a generalizable manner. Despite the absence of a formalized mathematical approach, the uncertainties should not halt our pursuit of research.
Statistics on Project Failures
Projects fail around 65% of the time.[3] While a disheartening statistic, it serves as a reminder that our resources can become losses if not handled wisely. The possibility that some actions might create more problems than they solve emphasizes the importance of accurately identifying methodologies and conceptual frameworks in AI research. Careful planning can avoid potential losses and unintended consequences.
The Promising Benefits of AI Technology
The potential benefits of AI technology are simply astounding. Utilizing AI to solve local productivity issues is a top priority, and it could be the catalyst we need. Innovating new ways to enhance the economy and personal lives can be substantial. However, realizing these benefits requires prioritizing alignment work. Ensuring rigorous consideration for human welfare before deploying technology is non-negotiable.
While the upsides of AI research are astonishing, the uncertainties and potential pitfalls demand a closer look. Let’s explore the vital need for caution, consideration, and alignment in AI research.
The Stakes of AI Research: A Call for Consideration and Caution
Acknowledging the Complexity: A Personal Reflection
This is my first time into pondering these concepts, I am compelled to tackle them with sincerity and diligence. The urgency of establishing a responsible research direction for AI, particularly for a vast and diverse demographic, is palpable. When executed well, the benefits can multiply exponentially.
The Opportunity and Challenge: A Thoughtful Approach to AI Research
Anyone who considers stepping into the world of AI research, Filipino or otherwise, must heed this reality: Understanding AI as a technology presents enormous opportunities. However, transforming such opportunities into triumphs rather than missteps demands more than mere instinct; it requires more:
Planning and Prudence: Embarking on this path warrants careful planning and utmost caution. The AI landscape, rich with potential, is equally laden with pitfalls that can result in profound hardship and sacrifice.
Learning from History: We must remember that uncharted domains of human endeavor often bring pain and struggle before innovation flourishes. Let our approach to AI be informed by these historical lessons, striving for thoughtful innovation rather than hasty experimentation.
A Responsibility to Proceed with Care
The stakes of our engagement with AI are indeed extraordinarily high. This is not just a technological venture; it’s a complex human enterprise that requires mindful consideration of ethics, alignment, and the social fabric of our communities.
I cannot envision AI research succeeding without individuals, teams, organizations, or governments factoring in alignment as a core component. AI systems think differently—using tokens, not words, and transformer models that is still not fully understood. Engaging in any research agenda without exercising the level of caution required by alignment is undoubtedly a recipe for disaster. A useful concept to illustrate this is whether an AI system exhibits corrigible properties. Is the AI system willing to allow itself to be modified or shut down when necessary? Even if the AI might not be directly utilized for output in research, it’s safer to recognize if these technologies are not yet robust enough to ensure safety. Taking the time to deeply understand how different AI systems are will certainly aid in advancing AI research.
We can build a sturdy bridge to AI success. We can reap the rewards of understanding AI technologies if we prioritize safety.
Contrasting Outcomes: Binisaya Buddy and Local Lore
I wrote a post on two striking examples that illustrate the importance of responsible AI development: Binisaya Buddy, a virtual guide reflecting Cebuano culture, values, and traditions, and Local Lore, a misguided attempt at cultural representation.[4]
Binisaya Buddy: Born in vibrant Cebu, this project represents the potential when alignment theory guides AI research. It serves both locals and tourists by authentically mirroring cultural nuances.
Local Lore: In contrast, Local Lore lacked local insight, leading to misrepresentations and disconnection from the community it aimed to serve. A poor execution created more problems than solutions.
Through the examples of Binisaya Buddy and Local Lore, we see both success and failure in aligning AI with cultural values. These case studies underscore the necessity of not only individual responsibility but also strategic collaboration at various levels to ensure that AI development is aligned with our cultural context and ethical values.
These contrasting outcomes highlight the importance of building an integrated system where various stakeholders work together towards responsible AI development. What form might such collaboration take, and how can it be cultivated to ensure that technology truly serves our community’s needs and values?
A Potential Pathway: Collaboration Between Government, Corporations and Academia
The journey toward harnessing the full potential of AI in the Philippines requires more than isolated efforts by individual sectors. It demands a cohesive and strategic collaboration between government bodies, corporations, and academic institutions. Here’s how can this tripartite collaboration can foster responsible AI development:
Government’s Role:
Regulatory Frameworks: By establishing clear and ethical guidelines for AI development, the government can ensure that technological advancements align with national values and societal needs.
Investment and Support: Government investment in AI research and development, infrastructure, and education can catalyze growth in the sector, supporting both public and private initiatives.
Corporate Sector’s Contribution:
Practical Application: Corporations leads in transforming theoretical research into practical solutions. Collaborating with academia to develop innovative technologies, industry players can focus on addressing local challenges.
Private Investments: By investing in AI and partnering with educational institutions, businesses can create a thriving ecosystem that promotes technological entrepreneurship and job creation.
Academic Institutions as Catalysts:
Research and Innovation: Universities and research institutions serve as the backbone of cutting-edge exploration in AI. Collaborative research with corporations ensures that academic findings translate into real-world applications.
Education and Training: Academia plays a vital role in nurturing the next generation of AI experts. By aligning curricula with industry needs and government visions, educational institutions can prepare a skilled workforce ready to contribute to responsible AI development.
Potential Cross-Sector Collaboration
Shared Goals and Strategies: By setting common objectives and collaborative platforms, the three sectors can align their efforts, maximizing efficiency and impact.
Ethics and Alignment: Ensuring that AI development adheres to ethical principles requires input from all three sectors. This collaboration ensures that technological advancements contribute to human flourishing without compromising cultural identity.
Conclusion
The possible collaboration between government, industry, and academia isn’t merely a beneficial alliance; it might be an essential strategy for responsible AI development in the Philippines. By recognizing their unique roles and working together towards a shared vision, these sectors can leverage AI’s transformative potential without losing sight of ethical considerations and local values.
The end goal is human flourishing.
Ultimately, the goal of any AI research, project, or even personal endeavor is human flourishing. To achieve this, a level of alignment work is required to ensure that the results will contribute to improving our lives. The larger the project, the greater the level of alignment work that must be included.
Our exploration of AI’s potential and challenges culminates in a broader perspective. With the lessons learned and the stakes understood, let’s envision the unique role that the Philippines can play in this rapidly evolving field, with an emphasis on alignment and cultural identity.
The Road Ahead: Philippines and AI
With 115 million inhabitants, including a vast English-speaking demographic, the Philippines stands at a crossroads. Our journey with AI is not merely about innovation; it’s about a responsible, ethical approach to technology. Both general AI research and alignment work are two sides of the same coin, vital for moving forward in an AI-shaped world. The stakes are immense, and the potential enormous. Our country must not settle for anything less than an approach that harmonizes general research with alignment to our shared values (global and local) and cultural identity.
Addendum
After finishing this write-up, I learned of a proposal by Congressman Mark Go, the Representative from the City of Baguio. He informed the Philippine Congress of his initiatives to propose a national drive to train 1% of Filipinos in AI technology and to establish a grant facility for AI research. Though no bill has been drafted yet, it goes to show that this write-up is both relevant and timely. You can see the presentation and with Doc’s commentary here.
Dominic Ligot is the Founder of CirroLytix, a social impact AI company, and Data Ethics PH, an online community focused on social issues such as data privacy, data security, AI-driven discrimination, data liabilities, data ownership rights, and data poverty. Three-time global winner of the NASA and ESA International Space Apps Challenges, his team’s award winning dengue surveillance application, AEDES, has been backed by the Group on Earth Observations (GEO), the Digital Public Goods Alliance (DPGA) and the UNICEF Innovation Fund.
When I refer to alignment research / work, I mean the effort that goes into understanding the alignment problem and its complexity. It’s hard for me to recommend alignment theory as a standard, especially since it’s not yet a concrete science on how to align AI systems and make them understand and preserve human values, and we’re all still figuring things out. However, as a reference, John Wensworth’s study guide is a good introduction to the complex nature of the theoretical areas intertwined with alignment.
But research shows that only 35% of the projects undertaken worldwide are successful—which means we’re wasting an extravagant amount of time, money, and opportunity.
Subtitle: Alignment and Misalignment: A Tale of Two AI Systems Bridging Culture and Technology in the Philippines
In the picturesque Island of Cebu, renowned for its vibrant heritage and breathtaking landscapes, an innovative AI system was built to bridge the gap between locals and tourists. Named “Binisaya Buddy,” after the local dialect, this AI system was designed with a deep understanding of Cebuano culture, values, and traditions.
Visitors arriving at the Mactan-Cebu International Airport were greeted by Binisaya Buddy through interactive kiosks. With a friendly virtual smile, the AI would share stories of Cebu’s rich history, explain the significance of traditional dances like the Sinulog, offer tips on enjoying local delicacies like Lechon, and even teach a few essential phrases in the local language.
But what set Binisaya Buddy apart was its alignment with the local culture. The AI was not just a repository of information; it was a reflection of the Cebuano soul. Locals were involved in its development, ensuring that it captured the nuances of their beliefs, values, and way of life.
Tourists found themselves immersed in an authentic Cebuano experience, facilitated by a technology that felt personal, respectful, and connected. Locals, on the other hand, felt a sense of pride in how their culture was represented and shared.
Binisaya Buddy became more than a tool; it became a symbol of how technology, when aligned with human values, can enhance cultural understanding, foster connections, and create meaningful experiences.
In another bustling city not far from Cebu, an ambitious project was launched to create an AI-driven tourist guide. Named “Local Lore,” this system was designed to offer tourists a glimpse into the local culture, traditions, and history.
However, unlike Binisaya Buddy, Local Lore was developed without local input, guided solely by algorithms fed with data from generic travel sites and social media. The lack of alignment with the community’s values and understanding of their unique cultural nuances led to unintended consequences.
Soon after its launch, locals began to notice misrepresentations. Local Lore described the Sinulog Festival as merely a colorful party, missing its deep religious significance. It recommended commercialized tourist traps over authentic local experiences, and its language translations were often inaccurate, leading to unintentional offenses.
The community felt alienated and misunderstood. They saw their rich culture reduced to stereotypes and misconceptions, a caricature rather than a celebration.
Tourists, initially excited by the novel technology, were left confused and disconnected from the real essence of the place. The lack of authenticity in the Local Lore’s guidance led to missed opportunities for genuine cultural immersion and understanding.
In time, Local Lore was abandoned, a failed experiment that served as a stark reminder of what can go wrong when technology is not aligned with the human values, context, and culture it aims to represent. It became a cautionary tale of how even well-intended innovation can lead to destruction when the alignment is overlooked.
Exploring the Responsible Path to AI Research in the Philippines
Introduction
The transformative potential of artificial intelligence (AI) is reshaping our world, presenting both thrilling opportunities and profound challenges. In the Philippines, where Dominic Ligot (Founder, AI expert & communicator)[1]and I have delved into AI’s possibilities, the integration of this technology into our unique cultural fabric can unleash exponential productivity. This post explores how we (as a country) can harness AI to enhance our local community without losing our distinct identity. With our shared fascination as a spark, I’ll examine the broader implications of AI, guided by the imperative to understand its complexities and the need to proceed with caution and adherence to alignment research[2].
Transformative Technologies and AI Research
The Transformative Potential of AI
AI stands as a transformative technology, holding the potential to reshape our world in the coming decade. The failure to understand and integrate AI could leave a country behind in a rapidly changing landscape.
Historical Precedence of Technological Transformations
Technological advancements have driven significant transformations. Whether through the application of cutting-edge physics in warfare or the discovery of new scientific fields, technological innovations have often led to pivotal moments in human history.
The Complexity and Urgency of AI Research
AI research is neither simple nor conceptually straightforward. The very complexity of the field underscores the importance of pursuing a deep understanding of it. While the path towards mastering AI may seem daunting and even wild, the stakes are too high for complacency.
The Inevitable Direction of AI Research
Our current situation leaves us with few choices but to embrace AI research. It’s not merely an option; it’s an essential direction we must take to ensure our survival in a future increasingly shaped by artificial intelligence. The question is not whether we should pursue AI research, but how we can do so most effectively and responsibly.
Having explored the immense transformative power of AI and its historical parallels, it’s essential to consider both the challenges and promises this complex field presents to us today.
Upsides and Downsides
AI Research Challenges and Uncertainties
AI research at present resembles alchemy more than mathematics. We still lack a formal theory on the malleability of data that can be infused into models, making our understanding of how data gets distributed to individual parts, such as tokens or layers, uncertain. These distributional shifts affect how technology can respond in a generalizable manner. Despite the absence of a formalized mathematical approach, the uncertainties should not halt our pursuit of research.
Statistics on Project Failures
Projects fail around 65% of the time.[3] While a disheartening statistic, it serves as a reminder that our resources can become losses if not handled wisely. The possibility that some actions might create more problems than they solve emphasizes the importance of accurately identifying methodologies and conceptual frameworks in AI research. Careful planning can avoid potential losses and unintended consequences.
The Promising Benefits of AI Technology
The potential benefits of AI technology are simply astounding. Utilizing AI to solve local productivity issues is a top priority, and it could be the catalyst we need. Innovating new ways to enhance the economy and personal lives can be substantial. However, realizing these benefits requires prioritizing alignment work. Ensuring rigorous consideration for human welfare before deploying technology is non-negotiable.
While the upsides of AI research are astonishing, the uncertainties and potential pitfalls demand a closer look. Let’s explore the vital need for caution, consideration, and alignment in AI research.
The Stakes of AI Research: A Call for Consideration and Caution
Acknowledging the Complexity: A Personal Reflection
This is my first time into pondering these concepts, I am compelled to tackle them with sincerity and diligence. The urgency of establishing a responsible research direction for AI, particularly for a vast and diverse demographic, is palpable. When executed well, the benefits can multiply exponentially.
The Opportunity and Challenge: A Thoughtful Approach to AI Research
Anyone who considers stepping into the world of AI research, Filipino or otherwise, must heed this reality: Understanding AI as a technology presents enormous opportunities. However, transforming such opportunities into triumphs rather than missteps demands more than mere instinct; it requires more:
Planning and Prudence: Embarking on this path warrants careful planning and utmost caution. The AI landscape, rich with potential, is equally laden with pitfalls that can result in profound hardship and sacrifice.
Learning from History: We must remember that uncharted domains of human endeavor often bring pain and struggle before innovation flourishes. Let our approach to AI be informed by these historical lessons, striving for thoughtful innovation rather than hasty experimentation.
A Responsibility to Proceed with Care
The stakes of our engagement with AI are indeed extraordinarily high. This is not just a technological venture; it’s a complex human enterprise that requires mindful consideration of ethics, alignment, and the social fabric of our communities.
I cannot envision AI research succeeding without individuals, teams, organizations, or governments factoring in alignment as a core component. AI systems think differently—using tokens, not words, and transformer models that is still not fully understood. Engaging in any research agenda without exercising the level of caution required by alignment is undoubtedly a recipe for disaster. A useful concept to illustrate this is whether an AI system exhibits corrigible properties. Is the AI system willing to allow itself to be modified or shut down when necessary? Even if the AI might not be directly utilized for output in research, it’s safer to recognize if these technologies are not yet robust enough to ensure safety. Taking the time to deeply understand how different AI systems are will certainly aid in advancing AI research.
We can build a sturdy bridge to AI success. We can reap the rewards of understanding AI technologies if we prioritize safety.
Contrasting Outcomes: Binisaya Buddy and Local Lore
I wrote a post on two striking examples that illustrate the importance of responsible AI development: Binisaya Buddy, a virtual guide reflecting Cebuano culture, values, and traditions, and Local Lore, a misguided attempt at cultural representation.[4]
Binisaya Buddy: Born in vibrant Cebu, this project represents the potential when alignment theory guides AI research. It serves both locals and tourists by authentically mirroring cultural nuances.
Local Lore: In contrast, Local Lore lacked local insight, leading to misrepresentations and disconnection from the community it aimed to serve. A poor execution created more problems than solutions.
Through the examples of Binisaya Buddy and Local Lore, we see both success and failure in aligning AI with cultural values. These case studies underscore the necessity of not only individual responsibility but also strategic collaboration at various levels to ensure that AI development is aligned with our cultural context and ethical values.
These contrasting outcomes highlight the importance of building an integrated system where various stakeholders work together towards responsible AI development. What form might such collaboration take, and how can it be cultivated to ensure that technology truly serves our community’s needs and values?
A Potential Pathway: Collaboration Between Government, Corporations and Academia
The journey toward harnessing the full potential of AI in the Philippines requires more than isolated efforts by individual sectors. It demands a cohesive and strategic collaboration between government bodies, corporations, and academic institutions. Here’s how can this tripartite collaboration can foster responsible AI development:
Government’s Role:
Regulatory Frameworks: By establishing clear and ethical guidelines for AI development, the government can ensure that technological advancements align with national values and societal needs.
Investment and Support: Government investment in AI research and development, infrastructure, and education can catalyze growth in the sector, supporting both public and private initiatives.
Corporate Sector’s Contribution:
Practical Application: Corporations leads in transforming theoretical research into practical solutions. Collaborating with academia to develop innovative technologies, industry players can focus on addressing local challenges.
Private Investments: By investing in AI and partnering with educational institutions, businesses can create a thriving ecosystem that promotes technological entrepreneurship and job creation.
Academic Institutions as Catalysts:
Research and Innovation: Universities and research institutions serve as the backbone of cutting-edge exploration in AI. Collaborative research with corporations ensures that academic findings translate into real-world applications.
Education and Training: Academia plays a vital role in nurturing the next generation of AI experts. By aligning curricula with industry needs and government visions, educational institutions can prepare a skilled workforce ready to contribute to responsible AI development.
Potential Cross-Sector Collaboration
Shared Goals and Strategies: By setting common objectives and collaborative platforms, the three sectors can align their efforts, maximizing efficiency and impact.
Ethics and Alignment: Ensuring that AI development adheres to ethical principles requires input from all three sectors. This collaboration ensures that technological advancements contribute to human flourishing without compromising cultural identity.
Conclusion
The possible collaboration between government, industry, and academia isn’t merely a beneficial alliance; it might be an essential strategy for responsible AI development in the Philippines. By recognizing their unique roles and working together towards a shared vision, these sectors can leverage AI’s transformative potential without losing sight of ethical considerations and local values.
The end goal is human flourishing.
Ultimately, the goal of any AI research, project, or even personal endeavor is human flourishing. To achieve this, a level of alignment work is required to ensure that the results will contribute to improving our lives. The larger the project, the greater the level of alignment work that must be included.
Our exploration of AI’s potential and challenges culminates in a broader perspective. With the lessons learned and the stakes understood, let’s envision the unique role that the Philippines can play in this rapidly evolving field, with an emphasis on alignment and cultural identity.
The Road Ahead: Philippines and AI
With 115 million inhabitants, including a vast English-speaking demographic, the Philippines stands at a crossroads. Our journey with AI is not merely about innovation; it’s about a responsible, ethical approach to technology. Both general AI research and alignment work are two sides of the same coin, vital for moving forward in an AI-shaped world. The stakes are immense, and the potential enormous. Our country must not settle for anything less than an approach that harmonizes general research with alignment to our shared values (global and local) and cultural identity.
Addendum
After finishing this write-up, I learned of a proposal by Congressman Mark Go, the Representative from the City of Baguio. He informed the Philippine Congress of his initiatives to propose a national drive to train 1% of Filipinos in AI technology and to establish a grant facility for AI research. Though no bill has been drafted yet, it goes to show that this write-up is both relevant and timely. You can see the presentation and with Doc’s commentary here.
Dominic Ligot is the Founder of CirroLytix, a social impact AI company, and Data Ethics PH, an online community focused on social issues such as data privacy, data security, AI-driven discrimination, data liabilities, data ownership rights, and data poverty. Three-time global winner of the NASA and ESA International Space Apps Challenges, his team’s award winning dengue surveillance application, AEDES, has been backed by the Group on Earth Observations (GEO), the Digital Public Goods Alliance (DPGA) and the UNICEF Innovation Fund.
You can find his complete profile here.
When I refer to alignment research / work, I mean the effort that goes into understanding the alignment problem and its complexity. It’s hard for me to recommend alignment theory as a standard, especially since it’s not yet a concrete science on how to align AI systems and make them understand and preserve human values, and we’re all still figuring things out. However, as a reference, John Wensworth’s study guide is a good introduction to the complex nature of the theoretical areas intertwined with alignment.
Quoted from The Project Economy Has Arrived by Antonio Nieto-Eodriguez of Harvard Business Review
From my blog: https://www.whitehatstoic.com:
Binisaya Buddy and Local Lore
Subtitle: Alignment and Misalignment: A Tale of Two AI Systems Bridging Culture and Technology in the Philippines
In the picturesque Island of Cebu, renowned for its vibrant heritage and breathtaking landscapes, an innovative AI system was built to bridge the gap between locals and tourists. Named “Binisaya Buddy,” after the local dialect, this AI system was designed with a deep understanding of Cebuano culture, values, and traditions.
Visitors arriving at the Mactan-Cebu International Airport were greeted by Binisaya Buddy through interactive kiosks. With a friendly virtual smile, the AI would share stories of Cebu’s rich history, explain the significance of traditional dances like the Sinulog, offer tips on enjoying local delicacies like Lechon, and even teach a few essential phrases in the local language.
But what set Binisaya Buddy apart was its alignment with the local culture. The AI was not just a repository of information; it was a reflection of the Cebuano soul. Locals were involved in its development, ensuring that it captured the nuances of their beliefs, values, and way of life.
Tourists found themselves immersed in an authentic Cebuano experience, facilitated by a technology that felt personal, respectful, and connected. Locals, on the other hand, felt a sense of pride in how their culture was represented and shared.
Binisaya Buddy became more than a tool; it became a symbol of how technology, when aligned with human values, can enhance cultural understanding, foster connections, and create meaningful experiences.
In another bustling city not far from Cebu, an ambitious project was launched to create an AI-driven tourist guide. Named “Local Lore,” this system was designed to offer tourists a glimpse into the local culture, traditions, and history.
However, unlike Binisaya Buddy, Local Lore was developed without local input, guided solely by algorithms fed with data from generic travel sites and social media. The lack of alignment with the community’s values and understanding of their unique cultural nuances led to unintended consequences.
Soon after its launch, locals began to notice misrepresentations. Local Lore described the Sinulog Festival as merely a colorful party, missing its deep religious significance. It recommended commercialized tourist traps over authentic local experiences, and its language translations were often inaccurate, leading to unintentional offenses.
The community felt alienated and misunderstood. They saw their rich culture reduced to stereotypes and misconceptions, a caricature rather than a celebration.
Tourists, initially excited by the novel technology, were left confused and disconnected from the real essence of the place. The lack of authenticity in the Local Lore’s guidance led to missed opportunities for genuine cultural immersion and understanding.
In time, Local Lore was abandoned, a failed experiment that served as a stark reminder of what can go wrong when technology is not aligned with the human values, context, and culture it aims to represent. It became a cautionary tale of how even well-intended innovation can lead to destruction when the alignment is overlooked.