![\"3D](\"https:\/\/www.centigrade.de\/wordpress\/wp-content\/uploads\/d686a520-717b-2ae9-440d-5168136f8be5.jpg\")
<\/h2>\n<\/h2>\nThe digitalization backlog in Germany seems to be slowly starting to clear up. Whether the reason for this is the 500-billion-euro special fund, the protection of our digital and analog sovereignty in times of war, or our dwindling competitiveness in the global market is, for now, irrelevant.<\/p>\n
<\/p>\n
The facts: things are changing.<\/strong><\/p>\n Some incorrigible optimists might even argue that waiting so long to make major investments could have been an advantage, because now we can learn from others\u2019 mistakes and do things \u201cbetter\u201d right from the start.<\/p>\n But do we really learn from others\u2019 mistakes? And can we really do things \u201cbetter\u201d?<\/strong><\/p>\n Learning from the mistakes of others first and foremost requires acknowledging that one is \u201cfallible,\u201d rather than assuming one was born with wisdom\u2014and, to the same extent, overconfidence\u2014spoon-fed. For some, this point alone is likely to be the first challenge.<\/p>\n Are there even any players who can learn from the mistakes of others? And if so\u2014who are they?<\/p>\n Only those with the financial leeway to do so can take action. It is therefore hardly surprising that the players tasked with resolving Germany\u2019s digitalization bottleneck tend to come not so much from traditional industry as from the fields of the military, cybersecurity, and GovTech.<\/p>\n And hovering over all these sectors\u2014as both savior and sword of Damocles\u2014is, of course, \u201cAI.\u201d<\/p>\n While the military and cybersecurity are essentially tasked with warding off \u201cthe evil,\u201d GovTech can at least claim to promote \u201cthe good.\u201d In this encouraging vein, the following three goals stand out particularly in GovTech:<\/p>\n In theory, this sounds like plenty of exciting work for digital designers. After all, designers are, by definition, people who can learn from the mistakes of others (and, incidentally, from their own) and who\u2014instead of acting like they know it all\u2014prefer to talk to people, understand their life contexts and needs, and use cost-effective prototypes to gradually arrive at the best possible digital solution for users.<\/p>\n No unnecessary features, no incomprehensible technical jargon, and certainly no misguided strategic priorities based on narcissistic overconfidence.<\/p>\n Digital designers excel at only one thing: systematically eliminating human IT conflicts.<\/strong><\/p>\n So if digital designers\u2014as digital design and UX professionals\u2014were the key players in implementing ambitious GovTech goals, if they held responsibility, had relevant decision-making authority, and could serve as strategic guides, then we would certainly be a significant step closer to the vision of doing everything \u201cbetter.\u201d<\/p>\n But unfortunately, this is yet another instance where utopia meets reality: If it was already a common mistake back then to make decisions based on overconfidence rather than on established human-centered design methods, why should that change today or in the future? The real point (or nonsense) of overconfidence is precisely that the overconfident person, even after making mistakes, still believes he or she acted correctly. And that\u2019s why:<\/p>\n The overconfident person is, unfortunately, generally not inclined to change anything or even promote someone else to their own decision-making position.<\/strong><\/p>\n And that brings us back to the well-known \u201ctoothless tiger\u201d model: We have plenty of highly trained digital design and UX professionals in Germany. Many of these well-trained individuals have engaged intensively with current topics that are essential for the GovTech sector. Topics such as accessibility and digital inclusion have been thoroughly explored and disseminated in communities, working groups, or at conferences like DDUX. One of the hopes of digital designers was that the Accessibility Enhancement Act (BFSG) would boost the chances of more user-friendly software in the public sector. Yet, as is well known, hope dies last:<\/p>\n To date, investments have not been made in the design of digital inclusion and accessibility, but rather in the implementation of purely technology-driven projects without concrete use cases or human-centered design.<\/strong><\/p>\n For example, while the EU research project iBorderCtrl was funded with public money, the call for proposals was purely security- and technology-driven, whereas accessibility and inclusive design did not stand out as core requirements\u2014and this despite the fact that a border process naturally affects people with very different abilities and limitations. Here, in particular, an accessible, low-stress, and transparent design would have been an obvious \u201cdefinition of done.\u201d<\/p>\n So large sums of money were poured into high-risk AI and algorithms\u2014but in this case, the question of what problem the AI was supposed to solve for the people affected, in what context, and with what value proposition seemed secondary at first\u2014just do it and see what happens.<\/p>\n The unfortunate consequence of a purely AI-focused approach is that digital designers are not in the positions where they could be the decision-makers they need to be.<\/p>\n So, more focus on AI and less on digital design? That\u2019s actually the wrong way to frame the issue. After all, human-centered design and AI are by no means mutually exclusive<\/strong>.<\/p>\n Any AI is only as good as the input it receives. It\u2019s no coincidence that the AI community is increasingly focused on the question: \u201cHow do you manage to provide a language model\u2014whether \u2018small\u2019 or \u2018large\u2019\u2014with the smallest possible context that still describes as precisely as possible exactly what the problem is that an AI agent is supposed to solve at a given moment?\u201d The systematic and technical answer to this question is now referred to as \u201ccontext engineering\u201d and makes it clear: \u201cIf you don\u2019t know what you want, the AI won\u2019t solve it for you either.\u201d That said, even the most ardent AI proponents now recognize that AI cannot deliver sufficient quality without the \u201chuman-in-the-loop.\u201d Without systematic human intervention, AI-driven solutions are neither scalable nor maintainable.<\/p>\n Anyone who uses AI must therefore know what they want. However, very few people\u2014especially those who overestimate their own abilities\u2014lack the knowledge of what they themselves want. But if one wants to successfully achieve the first of the three GovTech goals\u2014namely, improving citizen services\u2014it is unfortunately not at all about what one wants oneself, but rather about what others want\u2026 and are capable of. Not everyone has the same level of education; not everyone can see or hear.<\/p>\n New public service models that fail to take the diversity of our society into account in their details automatically relegate some users of public services to second-class status.<\/strong><\/p>\n Knowledge about end users\u2014their life contexts, limitations, and needs\u2014thus becomes more relevant when it comes to filling the context window of any kind of AI. This is especially true if an AI-enhanced public service is to be perceived by users as genuinely more user-friendly.<\/p>\n Building knowledge about the realities of life and the needs of end users and making it usable for AI thus becomes the key to better public services. It is about shifting perspectives and closing empathy gaps. Digital design and UX professionals can usually do this \u201cout of the box,\u201d but this expertise is of little use to them if they are not actively deployed in the key positions.<\/p>\n Yet it could be so simple: Digital design and UX professionals do what they do best: observe people, interview pilot users, define personas, map out user journeys, provide conceptual recommendations, and prototype and test existing solution ideas.<\/p>\n They feed this knowledge into a central AI-driven knowledge base in a systematic and structured form, and on the other hand, decision-makers (i.e., product owners or product managers) use this information via inspiring AI personas\u2014virtual users\u2014to generate ideas and prioritize backlogs. Even software engineers can benefit from this knowledge, as they can use the context established by digital design and UX professionals directly in their IDE via the Model Context Protocol (MCP). As a recent recording of the GermanUPA \u201cAI Working Group\u201d meeting impressively demonstrates, this vision is already being implemented step by step in an impressive manner.<\/p>\n If the GovTech sector wants to successfully achieve its primary goal\u2014namely, \u201cbetter citizen services\u201d\u2014this will only work if people and their needs are integrated into the process, and if digital design and UX professionals are brought in as the key stakeholders responsible for gathering this knowledge, translating it into concepts, and testing and evaluating current development stages with users.<\/p>\n Public tenders for technologies that impact people should, \u201cby design,\u201d also incorporate digital design and formulate concrete usability goals as the most relevant acceptance criteria. Funding must be allocated to areas where user research builds a connection with people and derives meaningful concepts from that connection. A prime example of how this could work is certainly the federal agency known as \u201cDigital Service\u201d. This digitalization partner of the administration currently has a human-centered team and, in particular, has advertised open positions in the field of user research. This gives hope that the funding allocated for this purpose will lead to significantly better results than in the past.<\/p>\n For even the most impressive endowment will come to nothing if it fails to address people\u2019s actual needs.<\/strong><\/p>\n And afterward, those in charge and the users will ask themselves: \u201cWhy did we fail to get it right again<\/strong>?\u201d Perhaps only the users will ask themselves this question, and not those in charge. Then this failure will become yet another stain on the long list of \u201clessons not<\/strong> learned.\u201d<\/p>\n","protected":false},"author":5,"featured_media":0,"template":"","tags":[840,934,479,1058,983],"class_list":["post-18175","blog","type-blog","status-publish","hentry","tag-ai","tag-digital-design-2","tag-digitalization","tag-govtech","tag-ki-2"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/blog\/18175","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/blog"}],"about":[{"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/types\/blog"}],"author":[{"embeddable":true,"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/users\/5"}],"version-history":[{"count":4,"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/blog\/18175\/revisions"}],"predecessor-version":[{"id":18180,"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/blog\/18175\/revisions\/18180"}],"wp:attachment":[{"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/media?parent=18175"}],"wp:term":[{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.centigrade.de\/en\/wp-json\/wp\/v2\/tags?post=18175"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}GovTech and Co.<\/h2>\n
\n
Good Times for Digital Design?<\/h2>\n
Utopia Meets Reality<\/h2>\n
Context is King<\/h2>\n
Seeing the world through different eyes<\/h2>\n
A possible path for the future<\/h2>\n
Conclusion<\/h2>\n