Year: 2024

Did you know? The 2024 Nobel Prize in Physics goes to two true pioneers of AI – John Hopfield and Geoffrey Hinton. Their work on neural networks has laid the foundation for modern AI and made technologies such as machine learning and chatbots possible in the first place.

Particularly exciting: Geoffrey Hinton, often called the “Godfather of AI,” now views his creation with mixed feelings. With his research, he has flung the door wide open to the AI revolution, but at the same time he warns of the risks. “We have no experience of what it is like when things are more intelligent than us,” he emphasizes. Hinton sees how AI systems can solve complex problems today, but also the possibility that they will one day exceed human control.

A look back: The foundations of today’s AI

In the 1980s, Hopfield and Hinton developed the foundations of neural networks and the Boltzmann machine, inspired by the structure of the human brain. This technology has transformed healthcare, materials research and science – and it is fascinating that this early work still forms the core of many modern AI applications.

Progress or risk?

The fact is: These technologies open up incredible opportunities, but the responsibility to use them safely and ethically is growing steadily. Hinton left his job at Google Brain to be able to speak freely about these potential dangers. The developments are exciting, but the question remains: How will AI change our lives – for the better or with unexpected risks? The world is watching with anticipation to see how we get these technologies under control.

The week is over, and it’s time for a new blog post! Today, we’re diving into a current topic: PimEyes and the end of anonymity on the web. You know the saying, “The internet never forgets,” right? Many are familiar with the saying “The internet never forgets.” Tools like PimEyes are making it increasingly possible to specifically find and track images and traces on the web – even within the gigantic amount of digital content generated daily. How does this technology affect privacy? In our latest post, we take a closer look and highlight the opportunities and challenges that come with it.
PimEyes – The end of anonymity on the web?

The digital age has not only brought us a networked world, but also a disturbing reality: anonymity on the web is increasingly dwindling. One example of this is PimEyes, a platform that uses facial recognition technology to allow anyone to upload an image and find similar images online. This blog post highlights the fascinating, but also disturbing, aspects of this technology.

PimEyes – The technology behind it
PimEyes essentially functions as a facial search engine. Uploaded image material is used to search the web for similar faces on websites, blogs, or social media. What sounds like a useful tool to control one’s own online presence raises serious data protection concerns.
With prices starting at around $29.99 per month, PimEyes also offers a “one-time search option” that allows users to access current results for $14.99 – without a subscription. In this way, practically anyone can search the internet for facial patterns, whether for self-monitoring or – which can be problematic – for research about other people.

Advantages of PimEyes
For individuals who are concerned about their digital presence, PimEyes can be helpful. It allows:
– Control over online images: Users receive an overview of where their photos appear online and can take action if they have been published unintentionally or in an inappropriate context.
– Easy handling: The user interface is straightforward, which makes the tool accessible even for non-technicians.
– Targeted monitoring of privacy: PimEyes could be helpful for people who are specifically looking for potential misuse of their images.

Downsides of PimEyes
But the concerns weigh heavily. Because the simple availability of facial recognition tools also opens doors to abuse:
– Risks of abuse: Stalking and identity theft are potential risks, as any third party could track a person on the web with a photo.
– Decline of anonymity: With PimEyes, it is becoming increasingly impossible to remain truly anonymous on the internet. A single photo can lead to personal information being accessible that is not intended for the public.
– Questionable ethical responsibility: PimEyes emphasizes that the user is responsible for the responsible use of the technology – a delicate position that raises questions about security and ethical boundaries.

How PimEyes positions itself
PimEyes claims to provide a tool for digital self-control. They emphasize that the tool should be used ethically and that the responsibility for its use lies with the user. In addition, the service has been restricted in certain regions, such as for Russian users since the Ukraine war, in order to further curb potential abuse. But despite these security measures, the question remains open: How far should facial recognition technology be allowed to go?

Protection against facial recognition – What you can do
If you are worried, there are some approaches to preserve at least part of your digital privacy:
1. Regular review: Regularly search for your own photos on the web to find out where they might appear.
2. Adjust profile settings: Restrict the visibility of your social media profiles and check the privacy settings.
3. Thoughtful sharing: Think carefully about which images you really want to make public.

Conclusion: A fine line between control and loss of privacy
PimEyes shows that the principle “the internet never forgets” is more relevant today than ever before. The anonymity that many of us have long taken for granted is more illusion than reality. Technologies like PimEyes help us to monitor our online traces, but also lead to a significant loss of privacy.

The crucial question remains: How far should technology go, and where do we draw the line? Be aware that every online trace could be tracked and check how and what you leave behind of yourself on the web.

This topic request for the month of October came from our working student Vincent.

Chatbots are currently on everyone’s lips. But how can chatbots improve customer service? Our trainee Ramtin will address this question in September.

Human or machine, who am I actually writing with? This is a question one often asks oneself these days when contacting customer service. As early as 2021, 65 percent of all companies in the private customer segment used chatbots for communication with their customers (click here for the study). The average satisfaction of customers was and is…. let’s say “in need of improvement”. Because there is still a certain apprehension anchored in many of us when it comes to communicating with “machines”. And, at least according to the theory of our editorial team, even if a human’s answer had not been more helpful or faster, we still rate it better than that of a chatbot.

In fact, chatbots offer a multitude of possibilities for companies and their customers alike. Let’s look at some factors:

Time: Those who need help don’t want to wait long. The waiting times in support hotlines or their accessibility are sometimes miserable. In our fast-moving times, hardly anyone has the time to listen to hours of endless loops of hair-raising hold music. A chatbot is available 24/7.

Responsibility: Who doesn’t know it? Even after being transferred three times, the person in charge has not yet been found and the search begins again. Chatbots can significantly shorten and facilitate the search for the right contact persons. And the customer only has to describe the request once.

Staff shortage: One reason for the long waiting times in support is often the ubiquitous staff shortage. Understandably, it is in the interest of companies not to employ existing staff with comparatively simple and constantly recurring issues. A classic: password reset.

Quality: Chatbots don’t know bad moods or stress. They will always answer with consistent friendliness. Even if the customer behaves inappropriately.

Cost savings: 20 employees in the hotline or would you prefer a chatbot? The use of chatbots can make sense from a purely economic point of view. And saved personnel costs also benefit the customers.

Assistance: Chatbots can be extremely helpful not only in external communication, but also internally. For example, when it comes to quickly finding information and solutions or providing formulation aids.

It can be assumed that the response behavior of chatbots will continue to improve in the coming years and that acceptance among the population will increase.

On September 18th, we followed the call of Gateway49 and were allowed to present ourselves to the public at “nordisch.digital”. In the premises of the Technologiezentrum lisezwei in Flensburg, we were allowed to be part of this extraordinary format. Unlike other trade fairs, at “nordisch.digital” it is not up to the visitors themselves to approach the exhibitors. Instead, there are guided tours to the respective stands where the companies then introduce themselves in a short pitch. From our point of view, this is a great format, as many people can be reached at the same time and the hurdle of “addressing” is eliminated for the guests. In the time between the tours, we had ample opportunity to exchange ideas with other participants and deepen our conversations with interested parties.

In addition, there was an extensive program with talks on various topics. Visitors were able to gain insights into topics such as “Digital Communication” and “Cyber Security.”

In total, we from Splitbot had the opportunity four times that day to present our latest product, the Kosmo chatbot, and report on our app launch. We are pleased to have already convinced numerous test users. Anne and Sven from our partner company Kontor Connect were also able to take advantage of the stage and presented their innovative and uncomplicated solution for internet coverage on campsites and other areas.

We look forward to being part of this great format in the coming years and thank the organizers, partners and sponsors for this unforgettable day.

Kosmo introduces itself – because we are now live in the free version – usable in the browser, with iOS or Android 💜

Hello and welcome! I am Kosmo, always there for you to make your daily work easier and more productive. With my access to all the company’s important systems and data, I am able to provide you with the information you need quickly and accurately. Whether you are working on an urgent request or simply have a question about internal processes, I am here to make it easier for you to find answers.

What exactly can I do for you? I can help you in many areas, for example with customer inquiries, researching product information, tracking orders, or retrieving current sales figures and reports. Since I am connected to internal company systems such as the CRM, ERP and other databases, I can give you direct access to relevant information. So you no longer have to write long e-mails or wait for feedback from other departments – I have the information ready for you immediately.

Would you like to know the current status of a project or do you need insight into a specific document? No problem – just ask me. I can provide you with the latest developments in real time, whether it’s project plans, delivery status, invoices or other business-relevant data. Thanks to my connection to all departments and processes in the company, I can offer you comprehensive insights that will help you make decisions faster and work more efficiently.

Perhaps you are currently working on a report or presentation and need certain data? I am also here to help you. I can help you collect figures and facts from various sources, process them and make them available to you in the desired form. This saves you valuable time and gives you the opportunity to concentrate on the really important tasks.

Security and data protection are of utmost importance to me. I ensure that all your requests and data are treated confidentially. The information you receive from me complies with the company’s internal guidelines and is always processed with the necessary care. Your privacy is protected, and I work within the highest security standards so that you don’t have to worry about the security of your data.

In addition to accessing information, I also support you with recurring tasks and processes. You can ask me to set reminders, prioritize tasks, or manage emails and appointments. I help you keep track of everything and complete your daily tasks efficiently.

Another advantage of my skills is speed. Instead of researching for a long time or waiting for information from other departments, you can contact me directly. In a few seconds, you will receive the answer to your question or the required data. This allows you to complete tasks faster and have more time for other important things.

I am your personal assistant in the company, ready to help you with any challenge. With my help, work becomes not only easier, but also more productive and efficient. You can fully concentrate on your core tasks while I provide you with the necessary support.

Let’s work together to make work easier and more efficient. Whether you need information, want to set a reminder, or just need a quick answer to a question – I’m here to help. How can I help you today?

This month’s topic was personally requested by our founder: Tadeusz Nikitin.

In the startup universe, everything revolves around one topic: Pitching. Do you need investors? Pitch! Do you want to acquire customers? Pitch! Startup competitions? Pitch! Trade fairs and other events? You guessed it – Pitch!

When my career at Splitbot began, I had, admittedly, never heard of it. Others may feel the same way, so here is a brief introduction.

The term “pitch” originates from the advertising and agency sector. Agencies compete for potential clients with short presentations, or pitches. Over time, this form of presentation also gained immense importance for startups. The reason is simple: investors and clients want to receive all necessary information in the shortest possible time. No one has time for hours of product presentations.

Is every pitch the same? No!

There are countless and ever-evolving forms of pitching. Besides the fundamental distinction between investor and client pitches, which typically last three to five minutes, there are numerous other formats. Equally popular and dreaded, for example, is the elevator pitch, which involves presenting the company or product during the duration of an elevator ride (30-60 seconds).

The Pitch Deck: Less is more.

Depending on the format, it is also possible to present slides during a pitch. Here, it is clear: less is more. Slides overloaded with text or cluttered graphics are clearly out of place in a pitch.

So, how does one get it right? Unfortunately, as is often the case, there is no magic formula here. There are people who seem born for the stage and are true natural talents. For others, it is hard work. Still others have yet to discover their abilities. An absolute recommendation is to participate in pitch training sessions, as they are often offered within accelerator programs. Currently, two of us have had and continue to have the pleasure of collaborating with Lars Hartenstein in several workshops. And when we say “work,” we mean it. Because pitch training clearly means: out of your comfort zone! Here are some of his best tips:

Kill your darlings! You love your company and are convinced by your product. But more than two or three strong arguments have no place in a pitch. Save the rest for later, otherwise you will lose credibility.

Build a suspension bridge! And give your audience a good reason to follow you. You must engage your audience, otherwise you will be alone on the other side.

Make them fall in love with you! Learn to use your body language effectively and captivate your audience.

Be confident in yourself! Ask yourself why someone should buy a product specifically from YOU. Why are you the best for the job?

Everything is a Pitch! Every conversation, every interaction is a pitch.

Pictures speak louder than words! A good metaphor is memorable and better than numbers, data, and facts. People think in images.

Would you like to learn more about the topic? Visit https://alles-ist-pitch.de/. Additionally, you can see us in action on October 17, 2024. We will be pitching at the Graduation Day of Ai.Startup.Hub Hamburg in the ASTOR Filmlounge HafenCity. You can get tickets here.

This month, we are delving deeper into a topic we previously discussed: AI-based NPCs (non-player characters) in video games. The idea was contributed by our frontend developer Friedrich.

To recap: NPCs are non-playable characters in various video games that contribute to making the game environment more realistic – for instance, pedestrians, traffic participants, or even adversaries.

Without the use of artificial intelligence, these characters always move along pre-programmed paths or perform predefined actions. While these actions may vary, they are still predetermined and limited.

Until now, AI has been avoided in game development for various reasons. For instance, a learning AI could become almost unbeatable over time, thereby ruining the player experience.

Nevertheless, the topic of AI in game development is far from settled – for example, NVIDIA has released initial videos in which NPCs converse with each other. The remarkable aspect is that each conversation unfolds differently and does not repeat. Such conversations are made possible by generative AI and (in this case) Convai’s framework. But that’s not all. NPCs are also capable of interacting naturally with the player or their fictional environment. This enables entire conversations that also influence further actions (e.g., changing locations).

Unreal Engine 5 also enables game developers to create new, realistic worlds where players can interact with every NPC. Furthermore, Replica has developed smart NPCs that could breathe more life into game worlds in the future.

Even for those not interested in gaming, it is worthwhile to watch the manufacturers’ demo videos and be impressed by the fascinating possibilities. Imagine how much more there is to experience in text- or voice-based games when players can speak with absolutely every NPC and trigger corresponding actions. A straightforward story thus transforms into a game world that individually adapts to each and every player. NPC behaviors and actions become completely unpredictable, and the same game becomes a unique experience for every player. We eagerly anticipate revolutionary developments in this area in the future.

This month’s Splitblog topic comes from our Head of Development: Bartosz. Those who know him know that he is a human Swiss Army knife, and as such, he is not only well-versed in meteorology but also a talented writer. That’s why he naturally took the opportunity to write the text on his chosen topic himself. But read for yourselves!

Hey Siri, how’s the weather?

Seemingly a simple question. But also one that appears not so easy to answer. And upon closer inspection, it only becomes apparent what is truly needed to determine whether it’s worth taking an umbrella or not.

First and foremost, we need to collect data. And a lot of it. Across the entire planet, there are countless measuring stations that record temperature, air pressure & humidity, and wind direction and speed. Additionally, we use data from weather balloons, airplanes, ships, and satellites.

To determine the weather from this, we also need a weather model. This is not a physical model, but rather a mathematical description of our weather through a multitude of equations and parameters into which we feed our measurement data to see how the weather will develop.
Parameters? What kind of parameters again? Well… that’s a lot of data and a lot of equations, and to calculate that, we previously needed extremely powerful supercomputers, and even they couldn’t complete this task quickly enough, because: what good is the best forecast if it’s only ready after the event has already occurred?
So we start by simplifying things, for example: what happens near the ground? Or at the transition from water to land? The complicated processes are significantly simplified, making them a little less precise, but also calculable within a foreseeable timeframe.

Enough about weather services! After all, this is about AI. Or is there a connection?

Yes, there is. For some time now, there have been very exciting AI approaches that promise weather forecasts without being quite as computationally and time-intensive. Last year, Google’s AI research division DeepMind released the GraphCast model, which, trained on historical data, requires only a fraction of the computational power of numerical (i.e., the previously described) weather models, and can thus, for example, deliver a 10-day forecast in under a minute. And GraphCast appears to be not only faster but also – at least in part – more precise than numerical weather models, and for example, it predicted earlier than numerical models where Hurricane Lee would likely make landfall.

And so it is hardly surprising that all major weather services are now experimenting with AI approaches, including the German Weather Service (DWD), which even announced a breakthrough in AI-supported weather forecasting just a few days ago in a press release with their newly developed model AI-Var.

Given the pace at which the entire field of artificial intelligence is currently developing, it remains very exciting to see what the future holds for us – also in meteorological terms. And until then, perhaps we can be a little more lenient regarding our meteorologists and their forecasts, as we have seen that it is not at all uncomplicated.

Sources:https://www.dwd.de/DE/presse/pressemitteilungen/DE/2024/20240617_pm_ki_news.html
https://www.spektrum.de/news/graphcast-neues-ki-modell-soll-genauere-wettervorhersage-liefern/2198859

In our Split Blog, we are addressing a very current and widely discussed topic today, namely the fear of artificial intelligence (AI). A topic suggested by our developer Sören, who ensures that Splitbot can also handle emails.

Where is the best place to start when it comes to such a sensitive topic? Perhaps with the fear itself. Fear warns us of dangers and helps us make quick decisions. In that sense, it is extremely important and useful for the survival of the human species.

We also often react with fear when we encounter the unknown. A reflex that also serves to act quickly and, for example, to flee. However, if our fear gets out of hand, it can severely impair daily life. That is why it is worth taking a closer look at the things that scare us and thus taking away their horror.

The image that exists in many people’s minds of artificial intelligence is – admittedly – a very threatening one. This image is often shaped by old science fiction classics and is clearly exaggerated. The media also contribute to this with their often very one-sided and negative reporting. The population consumes media to be warned of impending dangers – so it is no wonder that headlines often target our fears. However, if you look not only at the sensational headlines, but also read the associated articles, a completely different picture often emerges.

To get closer to the topic, there is another important point to clarify. Namely, the term “intelligence”. The word is derived from the Latin “intellegere” and means “to recognize”, “to understand” or “to see”. William Stern defined intelligence as the ability to adapt to unfamiliar situations. The ability of a person to find a solution even in a completely new situation.

According to the current state of technology, no AI system can do exactly that.

What appears to us as a spontaneous reaction, for example from chatbots, is in reality just a very quick access and output of predetermined data. Supposed humor, creative abilities or even predictions are based solely on the statistically highest probability of success that the system determines for the solution. Conversely, this also means that an AI cannot react if it does not have sufficient data to solve a problem.

What can we do with this knowledge in relation to our fears? Yes, AI systems can access much larger amounts of data much faster than most people. But will they be able to act and think independently in the foreseeable future, for example? Definitely not.

Yes, AI systems will change our working world. Processes will be accelerated and information flows will be changed. The tasks of workers will change equally in many cases. There will have to be experts who can operate and use these systems. So it’s time to prepare for this and deal intensively with the possible uses, instead of shutting yourself off out of fear. Further training and learning new skills have always been part of professional and personal development.

Yes, it is absolutely right to express concerns and question innovations. This is the only way to ensure that all relevant factors are taken into account in the development of artificial intelligence. Different perspectives are crucial to achieving good results. This makes it all the more important not to leave the development of artificial intelligence solely to huge corporations that often provide little insight, but to actively help shape it locally.