Design Community: Singaraja33

Humans and the retain of control in a world where AI thinks and decides alongside us

Singaraja33 — Sat, 04 Apr 2026 06:24:25 +0000

It's not the first time we write on this topic, but it's relevance makes it worth it because the evolution of AI as a whole could easily make it possible that in just a few months from now, you might be making an important decision and not remember if it was actually yours. And not because you forgot, but because the line between your thinking and the machine’s suggestion will simply have quietly disappeared.

That’s not something futuristic anymore, it’s already happening.

As mentioned in previous articles, we are entering a phase where artificial intelligence doesn’t just assist us, but participates in our processes, suggests us things and is able to refines, anticipate and sometimes even act. And while that sounds like progress, and in many ways it is, it raises a deeper question that most of us are only beginning to notice.

This whole thing is not just a technical problem but a purely philosophical one, a design challenge and ultimately a human one, because while for many years software followed a simple pattern of sending instructions with the matching executing them, that relationship has now changed.
Modern AI systems no longer wait for explicit commands and instead they anticipate intent, generate options and shape decisions before you even realize it. They act less like tools and more like collaborators. This shift is subtle, but it is one of the most important changes in the history of software.

Once a system begins to shape your options, it begins to shape your decisions, and when this happens, control is no longer about who clicks the button and it turns into something about who influenced what the button does.

In the midst of all this, still most people believe they are in control simply because they are the ones interacting with the system, but we must understand that control is not about interaction but about understanding and intention.
If a system suggests the best option, frames the problem and filters the available information, your role changes. You are no longer fully deciding, you are just confirming, and this is a totally different thing.

This basically creates an illusion of control. You feel in charge but the system has already narrowed the space of possibilities. You are choosing, but only within boundaries you did not define.
And don't get me wrong, because this is not necessarily harmful. In many cases it's actually incredibly useful, but it just changes the whole nature of decision making in a way that is easy to overlook.

Now consider what happens when something goes wrong...An AI system helps write production code, approve a financial decision or recommend a medical action. The outcome is flawed or harmful. At that point, a difficult question emerges. Who is responsible? Who can we go to blame??

Traditional systems of responsibility rely on clear agency and they are places where a person makes a decision, takes an action and is accountable for the result, but AI dramatically disrupts this clarity because now most decisions become the result of a mixture of human input, machine suggestion, training data and system design.
Responsibility does not disappear but it becomes distributed and it spreads across layers that are difficult to separate. And when responsibility becomes difficult to locate, accountability becomes weaker.

There is another bit change happening at the same time, one that is less visible but equally important, and this is that we are beginning to outsource not only tasks, but understanding itself. In our days it is increasingly common to accept generated code without even fully reading it, to rely on summaries instead of engaging with original sources and to trust explanations instead of building our own reasoning. This is efficient and of course often practical, but it introduces a quiet dependency that is very risky.

Over time, we begin to understand less about the systems we rely on, and despite sounding alarming we must also notice that this pattern has existed before. Look for example at calculators...When they appeared, they reduced the need for manual arithmetic. Also, GPS reduced the need for spatial navigation. There are other examples in the past, but the difference now is that AI operates at a higher cognitive level. It affects how we think, how we reason and how we make decisions.

If this trend continues without reflection, we risk becoming operators of systems we no longer truly understand. And of course nobody is saying we should be controlling every output or understanding every technical detail, that is no longer realistic, but instead meaningful control becomes something more practical and more necessary. We should keep an effort in recognizing when not to trust the system, understanding that blind trust is not control and can lead us to simply delegating without oversight.

Real control includes the ability to question outputs, to pause and to step outside the system when something feels wrong. It also means understanding the boundaries of the system. You do not need to know every parameter of a model but you should for sure have a sense of what it does well, where it tends to fail and what kind of information shapes its behavior. Without that awareness, the system becomes a black box that you depend on rather than a tool you use, and there is where the danger comes up.

And maybe the most important think to consider is that really meaningful control requires keeping the human intent at the center of the stage, prioritising and understanding that AI can optimize, suggest and automate, but it should not replace the underlying reason behind decisions. Humans should always be the ones defining goals, and systems should just be the excellent tools helping us to execute them. But when systems begin to influence or redefine those human goals, control starts to slip away.

There is a common idea in AI design that many might heard of and known as “human in the loop”. This idea suggests that as long as a human is involved in the process, everything remains under control. Nothing more far from the truth. In practice, this actually often becomes a simple formality where the system generates an output and the human approves it. But that is not meaningful oversightn but only passive validation.

True human involvement requires active engagement. It requires attention, critical thinking and the ability to intervene before outcomes are finalized. Without that, the human role becomes only symbolic rather than functional, and the machine will remain a defining factor.

In a world where almost anything can be generated, the real question is not whether something can be built but whether it should be used.

It is easy to think of this as a niche concern, something relevant only to developers or AI researchers, but that would be a big mistake because AI systems are already embedded in critical areas such as healthcare, finance, education or law. They influence decisions that affect real lives, and the way we design and interact with these systems will shape how responsibility, trust and authority function in society.

If meaningful control is lost, the consequences go beyond technical errors. They might affect accountability, decision making and the balance between efficiency and human values.

The solution is not simple and it's definitely not to reject AI or slow its progress. That is neither realistic nor necessary. Instead, and as we mentioned before, the real twist needs to happen in how we relate to these systems. This means questioning outputs instead of accepting them automatically. It means understanding systems well enough to recognize their limits. It means designing workflows where human reasoning remains central, even when machines handle most of the execution. And it also means accepting a new kind of responsibility, not just for what we directly create but for what we allow systems to create on our behalf.

In our opinion, the future of AI is not about machines suddenly taking control but more about humans gradually giving it away, with consciousness but often without noticing. Meaningful control does not and should not disappear all at once, but it should fade through convenience, efficiency and increasing trust in systems that seem to work most of the time.

If there is one thing clear is that we should ultimately preserve control, and maybe just redefine it in a way that actually fits the future we are building.

How Google is risking to lose a big chunk of the future tech race. Read it in Medium: https://luisyanguas22.medium.com/internet-is-changing-and-the-invincible-google-might-be-the-first-victim-76059f1582b2

Singaraja33 — Wed, 01 Apr 2026 04:39:51 +0000

luisyanguas22.medium.com

When Melania Trump walked the other day along a Figure AI humanoid... Read here 👇🏻👌🏻 https://future.forem.com/singarajatech/figure-ai-tesla-optimus-and-the-crazy-evolution-of-robotic-technologies-2hhn

Singaraja33 — Fri, 27 Mar 2026 07:08:39 +0000

Singaraja33

Mar 27

Figure AI, Tesla Optimus and the crazy evolution of robotic technologies.

#figureai #teslaoptimus #robotics #softwaredevelopment

Comments

5 min read

Figure AI, Tesla Optimus and the crazy evolution of robotic technologies.

Singaraja33 — Fri, 27 Mar 2026 07:05:18 +0000

When Melania Trump casually walked the other day into the White House alongside a humanoid 1,7m tall robot from Figure AI, many of you probable had this feeling that something surreal, almost crazy is going on. The machine, in a quite natural manner, just walked beside her, addressed world leaders in multiple languages and spoke as if it belonged there.

For a second, it felt like a scene pulled straight out of one of those science fiction movies from the 90's, but it wasn’t fiction and if that moment made you stop and think, it definitely should because it signals something much bigger than just a viral headline. In fact this simple moment might have marked the point where humanoid robotics stopped being experimental and started becoming part of our real world systems, education, labor and everyday life.

What we are witnessing right now with Figure AI and Tesla’s Optimus is not just technological progress but more of a big shift in how humans and machines will coexist. Until very recently, robots were either industrial arms locked in factories or machine looking prototypes struggling to walk without falling. They were specialized, limited and far from human like. We could not associate them to us at all, in any sense. But that’s no longer the case in a moment in time when the convergence of advanced hardware and breakthroughs in artificial intelligence, driven in part by organizations like OpenAI, has changed the trajectory entirely.

Figure AI’s humanoids, including the one that appeared at the White House, are designed to operate in human environments, and this is in fact a crucial distinction. Instead of building machines that require controlled conditions, they are building machines that are able to adapt to our world as it already exists. These robots can walk, manipulate objects, interpret instructions and, increasingly, communicate in natural language.

Tesla Optimus robot follows a similar philosophy but brings a different advantage: scale. To put it in context, Tesla is not just building robots, but it’s building them with the intention of mass production, and this is something Musk is constantly pointing out as anyone can see on his daily posts on X. Drawing from its expertise in manufacturing and AI systems used in autonomous driving, Tesla is aiming to make humanoid robots economically viable, and that’s the real unlock they seek because once these machines become affordable, their impact will not be limited to tech demos or high profile events and they will literally be everywhere.

What is actually incredible is not just their potential but what they can already do, as today’s humanoid robots can already handle logistics tasks like sorting, lifting and transporting objects. They can even assist in structured environments such as warehouses and factories where labor shortages are already a growing issue. They can also follow clear instructions, adjust to minor variations in tasks and operate for extended periods without getting tired at all.

And yet, this is still actually only the early stage, and the true real transformation lies in what comes next.

Imagine disaster zones with hundreds or thousands of harmed people where robots are first responders, navigating unstable structures and helping others with huge precision and without risking human lives. Think of hospitals where robots assist nurses by handling physically demanding tasks, allowing medical professionals to focus on care rather than logistics. Consider construction sites where machines take on repetitive or dangerous work, reducing injuries and increasing efficiency...
Even in education, the idea of robots becoming useful, something once dismissed as futuristic, is already being discussed at the highest levels.

The robot that walked with Melania was presented as part of a broader vision where humanoid systems could support personalized learning, potentially acting as adaptive, always-available tutors, and this is where the conversation shifts from “what robots can do” to “what role they will play in society”, because these machines are not just tools anymore and they are turning into pure collaborators.

That new panorama opens the door to enormous benefits in a field where labor intensive, repetitive and dangerous jobs could be offloaded to robots, reducing human exposure to risk. Productivity could increase dramatically, lowering costs and potentially reshaping entire industries. For individuals, despite of all the logical mental and real challenges, it could mean less time spent on physically exhausting work and more time focused on creative, strategic or meaningful activities.

But in the midst of all this new industry there a understandable tensions that can’t be ignored, because the exact same capabilities that make humanoid robots interesting and powerful also make them disruptive.
One of the biggest challenges many people talk about is reliability, because while human environments are messy, unpredictable and constantly changing, a robot might perform flawlessly in a controlled demo but struggle in a cluttered home or a chaotic construction site. And achieving true adaptability (the kind humans take for granted), is still an unsolved problem.

Cost is another barrier, because while companies like Tesla aim to reduce production costs, building a humanoid robot with advanced AI is still very expensive. For widespread adoption, these systems must become not just functional but economically practical, in a twist like the one that happened when wireless mobile phones appeared (initially extremely expensive and then becoming affordable over time)

Safety adds another layer of complexity, because when robots operate alongside humans, the margin for error becomes extremely small. A misinterpreted instruction or a mechanical failure could lead to real world consequences, and designing systems that are both capable and consistently safe is one of the hardest engineering challenges ahead.

Then there’s the question everyone is thinking about: jobs.
Automation has always reshaped labor markets, but humanoid robots take it further in many ways from the moment we understand that these machines are not limited to one function. They can, in theory, perform a wide range of physical tasks, and that raises concerns about displacement across industries, from logistics to retail or construction.

The transition won’t be simple and this is something even Musk admits. It will require mandatory factors like rethinking education, reskilling workers and creating new opportunities that align with a world where physical labor is increasingly automated.

And beyond economics, there is also something deeper, quite philosophical, as when a robot can walk beside a human, speak multiple languages or participate in public events, it starts to blur the line between machine and presence. It forces us to reconsider how we define intelligence, interaction and even companionship.

That moment at the White House was actually something quite symbolic. It showed that humanoid robots are no longer confined to labs or factories and they are entering public life, cultural spaces and global conversations, and as this has happened, there’s no easy going back.

The evolution of robotics is no longer a distant possibility and is unfolding in real time, faster than most people expected. Companies like Figure AI and Tesla are pushing the boundaries of what machines can do, but more importantly, they are redefining what machines actually are. We are moving towards a world where intelligence is not just something that exists in screens but something that walks, moves and interacts in physical space.

Robotics future has definitely arrived, and now it's time for us to see whether we’re ready for it or not.

[Meet Figure AI, the company hosted by Melania Trump]https://www.cnbc.com/2026/03/26/figure-ai-the-robotics-company-hosted-by-melania-trump.html

Translock IT
Author: Luis Carlos Yanguas Gómez de la Serna

Understanding the evolution and stabilization of the AI industry

Singaraja33 — Wed, 25 Mar 2026 06:08:53 +0000

If there is something we would all agree on is that the artificial intelligence landscape is evolving at a rhythm that very few of us could have predicted just a few years ago. From language models able of holding complex conversations to systems generating extremely realistic images and assisting in scientific research, the world of AI has moved from experimental laboratories to becoming an integral part of nearly everybody's daily life.

As we soon will enter April 2026, the industry is still characterized by rapid growth, huge competition and a mix of excitement and also uncertainty, but yet, signs are emerging that the sector may soon enter a phase of stabilization, offering clearer paths for businesses and everyday users alike.

Understanding how and why this transition might occur is crucial for anyone interested in the future of technology, and the first factor shaping this potential stabilization is basically technological consolidation. Over the past few years, an immense number of startups and research labs have launched their own AI models, leading to a fragmented landscape, and while diversity in all this fuels innovation, it also introduces challenges related to interoperability, quality control and standardization. In the midst of this crazy storm, users are logically asking themselves a number of questions.

Experts, on their side, suggest that the industry is likely to experience consolidation in the coming years, and this consolidation they think will mean larger companies either acquiring smaller players or collaborating within shared ecosystems. This process could well lead to standardized protocols for model deployment, greater compatibility across different platforms and more reliable user experiences, and by reducing fragmentation, AI will become more predictable and manageable, both for the businesses integrating these technologies and also for users relying on them for everyday tasks.

Efficiency and scalability represent another critical element in the evolution of AI because as of today, training "state of the art" models usually demand big computational resources, expensive hardware and significant energy consumption, and as these demands grow, we expect that companies will be incentivized to develop more efficient architectures and innovative training methods. Advances in model compression, optimized inference techniques and lighter architectures will allow AI systems to perform complex tasks while using less power and memory, and the result for this is an industry where AI becomes accessible not only in cloud based environments but also on personal devices like smartphones and laptops. Such accessibility not only helps broadening the user base but also stabilizes market expectations by reducing dependency on enormous data centers and costly infrastructure.

Regulatory frameworks will also play a central role in moving the AI industry toward stability, and we are sure that governments around the world will be increasingly aware of both the opportunities and the risks that advanced AI represent. Issues like privacy concerns, potential misuse and intellectual property have driven discussions on how to govern these technologies responsibly and this talks will only increase over time because a consistent and transparent regulatory environment is the only way for reducing uncertainty for companies, investors and consumers. Standards for transparency, safety and ethical deployment will also probably emerge, functioning similarly to the ISO standards in traditional industries, and with all those clear rules in place, businesses will be able to plan better their long term strategies without fear of sudden legal or social backlash, while users will gain trust in the safety and fairness of AI systems.

Equally important to the previous two is the professionalization of the sector. In the past, a substantial portion of AI development relied on enthusiasts and self taught talent, a kind of wild and unregulated field typical of many upcoming new undustries, and while this approach definitely accelerated innovation, it also introduced variability in quality and increased the risk of mistakes in deployment, and as the industry in itself matures, demand for highly trained professionals will only grow.

With regards to AI educational programs and certifications focused on AI safety and operational excellence, these are becoming standard, leading to a more qualified workforce that is not only better skilled but also aligned with industry best practices. This professionalization ensures that innovation continues without compromising reliability or accountability, further contributing to a more stable ecosystem.

Also, economic factors cannot be overlooked when talking about the sector's stabilization because AI remains an extremely capital intensive industry, with high upfront costs for research, model training and infrastructure. In the early stage AI market volatility, fueled by highly speculative investments and hype cycles, we saw a market that was certainly unpredictable at times, however, as consolidation and regulation take effect, business models are likely to become more predictable. Services based on subscription, enterprise licenses and targeted B2B solutions will for sure generate more consistent revenue streams for AI players, and the stabilization of these financial patterns will not only benefit investors but also allow smaller companies to participate in the ecosystem without risking crazy losses, fostering a balanced, sustainable industry landscape.

From the user standpoint, public perception and adoption also play a powerful role in stabilization, in a world where AI technologies are increasingly present in everyday life, from virtual assistants to tools supporting content creation, research and even revelant decision making. For the industry to really stabilize, users must strongly trust the systems they rely on, and for this to become a fact, transparency about how AI works, clear explanations of data usage and robust oversight mechanisms will reinforce public confidence.

As adoption grows alongside trust, the AI sector can also achieve a balance where innovation continues to grow but at the same time is guided by societal expectations and certain well understood standards. This alignment will ensure that technological progress does not outpace the social and cultural readiness to integrate AI responsibly.

Another emerging trend that definitely contributes to a more stable future is the coexistence of open source and proprietary models, a test that at times is quite challlenging. Open source AI initiatives push for experimentation, collaboration and customization, fostering innovation across the globe, but meanwhile, well funded proprietary models provide reliability, commercial support and optimized performance. The coexistence of these approaches becomes more and more necessary and also allows to create a diverse but structured ecosystem where individuals and organizations can choose solutions that match their needs while benefiting from established standards and safe deployment practices. The best of both worlds. This balance is fundamental and further reduces market volatility while encouraging sustainable growth.

In short, the stabilization of the AI model industry is likely to result from a combination of technological consolidation, improved efficiency, regulatory clarity, professionalization, economic predictability, public trust and the coexistence of open source and proprietary solutions. Each of these factors are key and address current uncertainties and risks, providing a framework for a mature, reliable and sustainable AI ecosystem that is the future both companies and users should aim for. And while innovation will remain a central driver, it should occur within boundaries that protect users and guide investment.

In view of all this things, and loking forward, the next few years are expected to solidify these trends. Most probably, companies will continue to push the limits of what AI can achieve, but they will do so in a more structured environment where expectations, responsibilities and outcomes are clearer.

On the other hand, users will interact with AI more confidently, investors will commit capital with higher predictability and governments will oversee the deployment of these technologies not in the unorganised or improvising way we sometimes see, but more responsibly. This convergence of technological, social, economic and regulatory forces will for sure mark the transition from the current period of rapid, sometimes chaotic growth to a more mature, stable phase where AI will become an integral, trusted and sustainable component of global society.

In essence, we don't expect that the future of AI models will be about slowing innovation but about channeling it in ways that create reliability, predictability and trust. By addressing fragmentation, inefficiency, legal uncertainty and public skepticism, the AI industry is positioned to move from the excitement of discovery to the steadiness of maturity, offering benefits that are both transformative and enduring, which is what everybody should be expecting.

The stabilization of this disruptive sector, maybe the largest tech change since the internet era, promises a world where AI is not just powerful but also properly integrated into the fabric of daily life, enhancing productivity, creativity and proper decision making across all areas of human endeavor. And the day we achieve this, the world will definitely be an even better place.

[The future of AI: How Artificial Intelligence will change the world]https://builtin.com/artificial-intelligence/artificial-intelligence-future

[The future of AI: Predictions for the next decade]https://www.gottabemobile.com/future-of-artificial-intelligence/

[The truth about AI regulation and the real winners behind it]https://luisyanguas22.medium.com/the-truth-about-ai-regulation-power-and-the-real-winners-behind-all-of-it-da9c0367c3d5

Author: Luis Carlos Yanguas Gomez de la Serna
www.translockit.com

Are we outsourcing our thinking to AI without noticing?

Singaraja33 — Wed, 18 Mar 2026 08:46:20 +0000

Involved as we are in this crazy hurricane of AI new developments, it's starting to be quite common to hear of situations where people are able to ship something in half the time thanks to AI, but then struggle to explain how it actually works.

This is something absolutely new, something that is slowly emerging around us as AI becomes part of everyday development work. We are entering a new tech chapter where teams can build faster than they can fully understand what they’re building, and while that is for sure powerful, it also carries a true risk that almost no one is talking about, because the reality is that we may be outsourcing not just effort, but understanding itself.

And this is not an argument against AI, in fact it might be quite the opposite. It’s an argument for using AI with more awareness, because the real opportunity right now is not just speed, but leverage without losing depth. This is where the real advantage comes from in our days.

Traditional software development has always followed a well known rhythm that was actually the root of true learning. You basically learned the fundamentals, fought with complexity, made mistakes and slowly built intuition. Understanding came from real experience and from being forced to figure things out the hard way.

AI changes that learning curve dramatically, because in our days a developer can simply generate production ready code in seconds, debug issues with guided suggestions or build systems they have never built before, creating an scenario where what used to take days or weeks can now take only a couple of hours.

This is a great and extraordinary progress, but it also introduces a challenge that is easy to overlook, when we start to realise that we are moving from doing the work to delegating the work. And delegation, by its own nature, creates distance.

The more we rely on AI to fill in the gaps and do it all, the less friction we experience, and while friction is often seen as something to eliminate, it is also where real understanding is formed. Without it, it becomes possible to build something that works without fully knowing why it works, and this is dangerous.

In the short term, this feels like a superpower and of course nobody can say it does not come with tons of advantages. Thanks to that, projects can move faster, teams can deliver way more and ideas turn into products at a speed that would have seemed unrealistic not long ago. But the risks that come with it cannot be ignored, and with all of this new fantastic trends and capabilities we are at the same time starting to see that sometimes teams are starting to experience simple and small bugs that take far longer than expected to diagnose, simple changes that create unexpected side effects, or small hesitations in the code that bring almost unpredictable effects.

These of course are not new problems for any developer, but they become more common when systems are assembled faster than they are understood. What is happening is simple but important. Execution is accelerating faster than understanding, and that gap matters more than it might seem at first.

In software, understanding is not optional because it's basically what allows you to debug with confidence, to scale systems reliably and to make proper architectural decisions. Without it you are not truly in control and you risk being operating something that works, but only within boundaries you do not fully see or reach to understand.

Every major abstraction in software has created a similar dynamic. We've seen similar things before. Frameworks, as an example, made development easier but hid complexity. Cloud platforms removed infrastructure burdens but introduced new layers that few people fully really understand, and tools without code allowed more people to build but sometimes at the cost of deeper understanding.

And while not being something new, AI is simply the most powerful abstraction we have ever introduced because maybe the big difference is the speed at which it is changing everything. What used to take years of gradual adaptation is now happening in months, and that compresses the time we have to rebuild understanding.

It would be easy to frame this as a problem, but that would miss the bigger opportunity. AI is not making people less capable but it is actually giving capable people unprecedented leverage, and developers and teams who benefit the most are not the ones who rely on AI blindly, but the ones who combine it with strong mental models.

When a developer or a team is able to understand the fundamentals, AI becomes a tool they can really direct, refine and challenge. They can spot when it is wrong, adapt its output to the desired context, and move faster without losing control.
In that sense, AI is an incredible tool to amplify capacities.

The real risk in this landscape is not the technology itself but how passively it can be used, and there is a huge difference between copying and pasting an answer and engaging with it. One approach creates dependency over time, while the other builds capability.

Active use of all the AI tools at hand today requires a bit of effort, but it strengthens understanding with every interaction, and over time, that difference becomes significant and creates a very big competitive advantage.

The most effective teams are not, and should not be stepping away from AI. They should be simply becoming more intentional in how they use it. They should treat it as a collaborator rather than an authority, always question its outputs, refining them and using them as a starting point rather than a final answer.
They also should continue to invest in fundamentals even when it might feel unnecessary in the short term. They should value the ability to explain a system as much as the ability to build it, as they always did in the past. In the end, they should create small habits that keep understanding alive, whether that means pausing to ask why something works or taking the time to rewrite a piece of logic to fully grasp it.

What we thing should come up from all this new wave is a new kind of developer and a new kind of team, not one defined by how much they use AI, but by how well they integrate it into their thinking, and we think the future of our industry will not be shaped by those who use AI the most, but by those who use it with intention and understanding.

AI is one of the most powerful tools ever introduced into software development. It can accelerate learning, reduce problema and unlock levels of creativity and productivity that were previously out of reach, but as we said before, its real value depends on how we choose to use it.

If we allow AI to fully replace our thinking, we lose something essential. If we use it to enhance our thinking, we gain an advantage that compounds over time. As simple as that.
Because in the end, the real advantage is not just building faster. It is understanding what you have built and why it works.

Creating an amazing son with AI is already possible and here is why! 👇🏻 https://dev.to/singaraja33/how-ai-is-rewriting-the-rules-of-music-creation-and-production-51p7

Singaraja33 — Sun, 15 Mar 2026 02:53:00 +0000

How AI is rewriting the rules of music creation and production

Singaraja33 ・ Mar 15

#musicai #softwaredevelopment #ai

Si quieres construir tu App tu mismo, rápido y prácticamente gratis, tienes que leer esto!! 👇🏻 https://open.substack.com/pub/luisyanguas/p/lovable-la-startup-de-ia-con-la-que?utm_source=share&utm_medium=android&r=7es6ar

Singaraja33 — Fri, 13 Mar 2026 04:24:30 +0000

Lovable: la startup de IA con la que puedes crear apps hablando con una máquina

La historia de cómo un ingeniero sueco está haciendo que cualquiera de nosotros pueda crear software mediante IA

luisyanguas.substack.com

Y si toda tu empresa funcionara gracias a agentes de IA?

Singaraja33 — Thu, 12 Mar 2026 05:26:40 +0000

Independientemente de que vivamos inmersos en un mundo que nos acostumbra a avances cada vez más impresionantes en lo que a tecnología se refiere, hace un par de semanas tuvimos una conversación de esas que te hacen pensar más tiempo del que esperabas.

Un amigo y posible cliente asiático con negocios en el sector logístico comentaba una idea cuanto menos ambiciosa. Pretendía construir una empresa logística de envíos internacionales de paquetería, pero diseñando toda la operación completa, de cero a cien, alrededor de diferentes agentes, o lo que se llama en inglés "agentic workflows". Es decir, tenía la idea en mente de crear un negocio integral basado solamente en flujos de trabajo gestionados por agentes de inteligencia artificial. Casi sin personas involucradas. Casi todo absolutamente automatizado.

No hablaba de añadir un poco de IA aquí o allá, ni de automatizar algunos procesos aislados, sino que su propuesta iba muchísimo más allá e implicaba que las operaciones centrales de la propia empresa estuvieran gestionadas por una red coordinada de agentes de IA.

A primera vista, y a pesar de ser conscientes de que sin duda avanzamos hacia esa dirección, ver este asunto planteado abiertamente por primera vez sonaba quizá demasiado futurista y poco realista en la práctica.

Quienes estén dentro del sector de la logística internacional, sabrán que es de por si caótica por naturaleza. Los paquetes muchas veces se retrasan, las direcciones pueden estar mal escritas, los documentos de aduanas fallan, los vuelos cambian de horario, y mil cosas más. Es un mundo lleno de excepciones y de temas que tradicionalmente siempre han necesitado participación humana.

Independientemente del reto, esa tarde hablamos con un desarrollador muy sólido a quien conocemos de hace años y que está especializándose cada vez más en modelaje de IA. Sorprendentemente, al profundizar con el en esta posibilidad, la idea tomó algo más de sentido. Y no solo eso, sino que él defendía que es precisamente la logística internacional uno de los sectores más adecuados para aplicar sistemas "agentic". Nos decía que era perfectamente posible plantear una arquitectura que permita operar gran parte del negocio mediante agentes de IA coordinados, manteniendo a las personas exclusivamente en aquellos puntos donde su criterio realmente aporta valor.

Evitando mencionar detalles específicos del proyecto, por motivos comprensibles, sí merece la pena explicar el enfoque conceptual que discutimos, ya que ilustra algo que mucha gente todavía no termina de entender sobre los agentic workflows. Y es que los sistemas agentic no tratan simplemente de sustituir trabajo humano sino que lo que persiguen es rediseñar cómo funciona una empresa, y cuando empiezas a pensar en un negocio desde ese ángulo, aparecen estructuras completamente nuevas.

Hoy en día, la mayoría de empresas logísticas serias tienen incorporados softwares que mejoran bastantes dinámicas de negocio, pero siguen siendo sistemas de perfil bastante tradicional. Estas empresas suelen contar con un ERP gestiona los pedidos, un software que ayuda a calcular rutas, un sistema de tracking que sigue los paquetes a veces incluso en tiempo real, o una plataforma de atención al cliente que gestiona consultas.
Cada uno de estos sistemas cumple su función, pero el problema de fondo es que no "piensan" juntos y esto provoca que, al final, las personas sigan siendo el verdadero pegamento que conecta todo y hace que la empresa funcione con sentido.

Un envío típico suele seguir una cadena bastante manual. Básicamente, el cliente introduce una solicitud, un operador revisa los datos y otro comprueba las rutas disponibles y selecciona un transportista. Se preparan los documentos, se pone en marcha el envío y si algo sale mal, otro equipo investiga el problema. Si el cliente necesita algo o tiene un problema, casi siempre también hay gente en atención al cliente que responde preguntas sobre el estado del paquete.

Este sistema de funcionamiento describe un sistema de diferentes células de negocio que principalmente lo que hace es ejecutar tareas, pero que no entienden el proceso completo, y eso genera tres fricciones constantes:

1- Una carga enorme de coordinación humana
2- Una reacción lenta ante problemas
3- Una inteligencia muy limitada entre sistemas.

Los agentic workflows cambian esta estructura por completo, y ahí está la gran revolución en todo esto, porque en lugar de que las personas coordinen herramientas, básicamente se puede diseñar una red de agentes que se coordinan entre ellos y que todos sepan todo de todo lo que está pasando.

Y no se trata de simples chatbots algo más avanzados que los de hace un par de años y con acceso a APIs, sino que se trata de herramientas muy potentes que se parecen más a un trabajador digital con un rol concreto dentro de la organización, y donde cada agente está interconectado y tiene responsabilidades claras, acceso a ciertas herramientas y la capacidad de interactuar con otros agentes de forma increíblemente eficaz cuando necesita resolver un problema.

En vez de construir módulos de software aislados, lo que se diseña aquí son roles operativos, y curiosamente este enfoque es donde desde el punto de vista técnico encaja muy bien con cómo funciona una empresa logística.

Durante nuestra conversación describimos una arquitectura en la que la empresa de paquetería estaría gestionada por un conjunto de agentes especializados. No un único modelo, o una única super IA, sino algo más parecido a un equipo de operaciones digital.
El punto de partida sería un agente encargado de la admisión de envíos y el trabajo de este agente sería básicamente validar nuevas solicitudes, comprobar direcciones, detectar posibles restricciones de contenido, revisar requisitos de aduanas y evaluar transportistas posibles.
En la mayoría de casos, este agente debería ser capaz de validar los envíos automáticamente en cuestión de segundos, y si dicho agente encontrará algo ambiguo o sospechoso, el sistema escalaría el caso a una persona de carne y hueso.

Una vez validado el envío, de manera casi siempre automática, pasaríamos a la fase del routing. En logística internacional, elegir la ruta adecuada es clave y suele significar equilibrar coste, tiempo de tránsito, disponibilidad de transportistas y restricciones aduaneras. Aquí, los sistemas tradicionales utilizan reglas fijas o algoritmos relativamente simples, mientras que un agente de routing puede razonar sobre estas variables de forma mucho más flexible, comparar múltiples APIs de transportistas, simular tiempos de entrega y adaptarse a cambios o interrupciones.
En lugar de calcular una ruta estática, el sistema toma decisiones operativas dinámicas.

Otra parte clave del sistema tiene que ver con lo que se conoce como "excepciones". En logística, lo normal suele ser lo fácil, pero a veces el verdadero trabajo empieza cuando algo se desvía del plan original, y eso ocurre por ejemplo cuando un paquete se queda bloqueado en aduanas, un vuelo se cancela, o un transportista local falla en el escaneo de un paquete concreto.

En las empresas tradicionales suele haber grandes equipos operativos que dedican su tiempo a investigar y resolver todos estos incidentes, pero otra de las enormes mejoras es que en un sistema agentic como el planteado, se puede añadir un agente de gestión de excepciones que monitorice continuamente los datos de tracking y los eventos de los transportistas. Cuando esté agente detecta algo fuera de lo normal, investiga el motivo, recopila información relevante y propone acciones correctivas. El resultado es que en la enorme mayoría de casos, este agente puede resolver el problema de forma automatizada, y solo los casos realmente complejos llegan a manos humanas.

Otra de las cosas que cambia radicalmente con este sistema de agentes es la comunicación con el cliente, en un sector y negocio donde las empresas reciben enormes volúmenes de consultas preguntando por el estado de los envíos, por problemas o por retrasos. La mayoría de estas preguntas siguen patrones bastante repetitivos, y un agente de comunicación conectado al sistema de envíos puede ofrecer respuestas muy contextualizadas y, lo más interesante, anticiparse a los problemas. Si el sistema detecta que un paquete probablemente sufrirá retrasos por líos en aduanas, puede avisar al cliente antes incluso de que éste pregunte, y ese pequeño detalle cambia por completo la experiencia del usuario.

Básicamente, en todo este mundo de agentes interconectados, lo que de verdad hace la magia es esa palabra tan escuchada últimamente y que es la famosa orquestación. Los agentes individuales son útiles, pero la verdadera potencia aparece cuando empiezan a coordinarse y "orquestarse".
En ese sentido, la arquitectura que planteamos incluiría una capa de orquestación que funcionaría casi como un sistema operativo para trabajadores digitales. El objetivo de esta capa sería gestionar cómo se reparten las tareas, cómo se comparte la información y cuándo es necesario que una persona intervenga.

Cuando un envío entra en el sistema, el agente de admisión puede activar al agente de routing. Este puede necesitar información del agente de documentación. A su vez, el agente de comunicación puede informar al cliente en distintos momentos del proceso, y así hasta el final.
La diferencia con los workflows tradicionales es que aquí el flujo no es rígido sino que los agentes pueden interactuar entre sí de forma dinámica, pedir ayuda, repetir operaciones o escalar decisiones, y en última instancia trasladar temas a un humano. Esa flexibilidad es lo que permite manejar la imprevisibilidad del mundo real.

Cuanto más analizábamos el concepto, más claro resultaba que la logística podría ser uno de los sectores donde los sistemas agentic despeguen de verdad. Por un lado, porque los procesos logísticos ya están organizados en tareas muy definidas, y los sistemas agentic simplemente digitalizan esos tareas o esos roles. Y por otro lado, porque el propio ecosistema logístico es de por si muy rico en APIs, donde los transportistas disponen de interfaces para reservar envíos y ver el tracking, los sistemas de documentación permiten integraciones automáticas, etc.
Todos estos hechos crean el caldo de cultivo perfecto para que los sistemas automatizados de agentes puedan entrar a trabajar.

Pero a pesar de todo ello, y esto es importante señalarlo, una de las cosas que también nos quedó clara es que este modelo de agentes no elimina a las personas sino que estas siguen siendo esenciales para manejar bastantes temas complejos dentro de la mecánica del negocio. Las máquinas, de momento, no cubren la capacidad humana de interpretar normativas, negociar acuerdos con transportistas o supervisar el sistema.

Lo que cambia básicamente es el nivel y la forma en que trabajan estas personas, que en lugar de procesar manualmente miles de envíos como siempre se había hecho, pasan por ejemplo a supervisar la operación automatizada y actúan solamente cuando aparece un caso realmente complejo.

Es un cambio parecido al de un controlador aéreo donde el controlador no pilota cada avión, pero supervisa el sistema entero.

Quizá la lección más interesante de esta oportunidad no tenía que ver solo con la logística internacional sino que tenía que ver más con cómo podrían organizarse las empresas en el futuro. En general, no solamente en esta industria, porque los agentic workflows hacen posible el diseño de organizaciones enteras alrededor de ciclos de decisión y ejecución autónoma, en lugar de depender de departamentos rígidos y software desconectado. Y esto vale para casi cualquier modelo de negocio.

Lo que sí tenemos claro es que, en adelante, algunas empresas simplemente añadirán funciones de IA a sus sistemas actuales. Otras irán más lejos y diseñarán operaciones completamente nativas en IA desde el principio. Y la diferencia entre esos dos enfoques probablemente definirá la próxima generación de startups que veremos en los próximos 1-2 años.

[]https://luisyanguas22.medium.com/meet-your-new-coding-buddy-ai-agents-are-taking-over-the-boring-stuff-7fef6395dc33
Translock IT
Luis Carlos Yanguas

Un nuevo enfoque en software, donde el AI-first acapara todas las conversaciones

Singaraja33 — Fri, 06 Mar 2026 05:26:45 +0000

A día de hoy, casi cualquier conversación sobre el desarrollo de un nuevo producto digital empieza con el mismo tema: la inteligencia artificial.

Cuando las empresas estudian hoy la posibilidad de construir una nueva plataforma, una app o un sistema interno para ellas mismas, el punto de partida ya no suele ser solo la arquitectura del producto o las funcionalidades básicas, todo aquello que hasta hace poco tiempo era básicamente la "materia prima" de cualquier proyecto. La primera cuestión que se pone hoy sobre la mesa es la forma en la que la IA podría intervenir en el proyecto en cuestión, y qué papel puede desempeñar dentro del sistema que se quiere poner en marcha.

Esta forma de enfocar los proyectos tecnológicos se está extendiendo con rapidez en cualquier mercado, y es especialmente notable en Asia, un mercado con capas de población masivas y mucho menos maduras que en occidente, donde muchas compañías están replanteándose cómo diseñar sus herramientas digitales para venderlas en un mercado que tiene la enorme ventaja de que decenas de millones de personas están viviendo toda esta revolución de la IA como una primera introducción al mundo tecnológico.

En general, las empresas de hoy buscan algo más que software funcional. Lo que quieren son sistemas capaces de interpretar información, automatizar procesos y ofrecer conocimiento útil a partir de los datos. Su objetivo ya no es solo construir software sino hacer sistemas inteligentes.

Los desarrolladores y empresas de hoy se han dado cuenta de que la IA ha dejado de ser una tecnología casi abstracta o simplemente una especie de maquina que responde preguntas. Son absolutamente conscientes de que la IA es algo que simple y llanamente forma parte del día a día de sus propias vidas personales y profesionales. Directivos, analistas, equipos de marketing o responsables de producto utilizan herramientas basadas en IA de forma habitual, redactan textos asistidos por modelos generativos, analizan datos mediante plataformas inteligentes o interactúan con interfaces conversacionales capaces de responder preguntas complejas.

En este contexto, es inevitable que cualquier empresario o emprendedor empiece a preguntarse cómo trasladar todo esto a sus propios productos y procesos, porque si una herramienta externa puede generar análisis en segundos o sintetizar grandes volúmenes de información, por qué no iba a poder hacerlo también en el propio software interno de la empresa?

A partir de ese momento, las conversaciones sobre tecnología con cualquier cliente empiezan a cambiar.
Los proyectos ya no se plantean solo como plataformas que organizan información, facilitan determinadas tareas o tienen X o Y funcionalidades, sino que hoy lo que la gente busca son basicamente sistemas que ayuden a comprender mejor lo que ocurre dentro de su organización y que contribuyan activamente a la toma de decisiones, agilizando y mejorando su eficiencia a casi todos los niveles y, por supuesto, abaratando costes.

El emprendedor de hoy sabe que la IA permite que el software no se limite a almacenar o mostrar datos, sino que pueda analizarlos, interpretarlos y transformarlos en conocimiento útil con una habilidad infinitamente mayor de la que recordamos en los inicios del machine learning, y esto abre la puerta gigantesca a nuevas posibilidades. Una plataforma ya no es algo casi estático, sino que con el uso de IA puede ser capaz de revisar automáticamente documentos y extraer información relevante. Un sistema interno puede identificar automáticamente patrones en los datos operativos y proponer oportunidades de mejora. Un entorno de trabajo digital puede ofrecer respuestas instantáneas a preguntas que antes requerían consultar manuales, informes o bases de conocimiento...

Cuando todas estas capacidades se integran de forma adecuada, el papel del software cambia totalmente y deja de ser solamente una herramienta de gestión, pasando a convertirse en un apoyo activo y real para el trabajo diario.

Otro factor importante para cualquier cliente es que son plenamente conscientes del crecimiento constante de los datos que generan sus empresas, y estos datos son factores importantes como sus propias interacciones con clientes, operaciones internas, etc...Toda está información y todos estos datos generan un potencial de negocio fortísimo que gestionado manualmente aporta poco valor, pero que operado y procesado mediante sistemas de IA se convierte en una herramienta clave a la hora de gestionar bien cualquier negocio.

Además, la inteligencia artificial no solo automatiza todo esto, sino que sus sistemas pueden detectar tendencias, encontrar correlaciones o identificar anomalías con una velocidad y una escala que sería imposible replicar manualmente, conviertiendo a la IA en un complemento muy potente para cualquier plataforma y empresa que gestione información compleja.

Para los desarrolladores de software de hoy, este nuevo escenario obliga a cambiar el foco en la forma de entender los proyectos tecnológicos. Ya no se trata solo de diseñar interfaces, construir APIs o garantizar la escalabilidad de una infraestructura, sino que es fundamental pensar en cómo fluye la información dentro del sistema y en qué puntos la inteligencia puede aportar verdadero valor.

Esto no significa que todos los productos deban girar completamente en torno a la IA. De hecho, una de las decisiones más importantes en cualquier proyecto, y algo de lo que ya hemos hablado ampliamente en algún otro artículo, es identificar cuándo y cómo tiene sentido utilizar la IA, y cuándo no.

Por ejemplo, en procesos que implican grandes volúmenes de información, tareas repetitivas de clasificación o análisis, o situaciones donde es útil identificar patrones complejos, la IA sin duda marca un beneficio evidente. En contextos de ese tipo, integrar modelos de inteligencia artificial puede transformar radicalmente la utilidad del producto, pero sin embargo algunos otros sistemas funcionan perfectamente con arquitecturas de software más tradicionales. Ejemplos de ello son por ejemplo procesos simples, flujos de trabajo más rígidos y bien definidos o aplicaciones centradas en transacciones que suelen requerir ante todo fiabilidad, claridad y robustez, pero no tanta interpretación.

El reto consiste en encontrar el equilibrio adecuado entre ambos mundos, en un sector donde cada vez es más habitual ver plataformas que combinan una base sólida de ingeniería de software con capas de inteligencia que amplían sus capacidades. En estos casos, la inteligencia artificial no sustituye al sistema, sino que lo potencia.

De cualquier forma, hay todavía muchos ejemplos, como es el caso de algunos marketplaces, donde la plataforma en sí misma sigue siendo el elemento central y donde es la propia plataforma la que proporciona la estructura, los flujos de trabajo y la experiencia de usuario, siendo después la IA quien actúa como una capa adicional multiplicando el valor del conjunto.

Esta evolución en la manera de enfocar proyectos también está cambiando la forma en que se plantean las primeras fases de los mismos. Como indicábamos al inicio, hoy, casi siempre, en lugar de empezar directamente por las pantallas o las funcionalidades, muchas veces se empieza intentando comprender mejor la información que maneja el cliente en cuestión y los procesos donde surgen más fricciones. Se dedica gran parte del tiempo inicial a analizar cómo circula el conocimiento dentro de una empresa, dónde se concentran las decisiones más complicadas o qué tareas consumen más tiempo. Y es a partir de ahí cuando se deciden los objetivos reales y cuando el diseño de la plataforma se estructura en torno a lo que pueda aportar la IA.

Mediante este método creativo, las plataformas digitales ya no se enfocan ni se valoran unicamente en base a las funcionalidades que puedan ofrecer, sino también por su capacidad para real de ayudar a las personas a entender mejor la información con la que trabajan, ampliando al máximo las expectativas sobre lo que cualquier producto digital puede hacer, en un nuevo enfoque tecnológico donde sin duda ya nada es como antes.

Habrá trabajo para el desarrollador futuro en la era de la IA? Este es un buen análisis: https://open.substack.com/pub/singaraja33/p/el-gran-reto-para-los-desarrolladores?utm_source=share&utm_medium=android&r=7es6ar

Singaraja33 — Mon, 02 Mar 2026 05:40:16 +0000

El gran reto para los desarrolladores en un mundo dominado por la Inteligencia Artificial.

Cómo adaptarse, evolucionar y prosperar en un mundo en el que la IA ya está empezando a reemplazar tareas muy importantes.

singaraja33.substack.com

The challenges for Europe in the AI battle https://luisyanguas22.medium.com/the-future-of-ai-is-being-built-now-and-europe-is-at-risk-of-missing-it-7cdcee806a8c

Singaraja33 — Sat, 28 Feb 2026 08:14:29 +0000

luisyanguas22.medium.com