top of page
Suche
Alexander K. Ammer

Future work: Human & AI teams


Dall-E generated picture with two persons with laptop looking into a data web flowing in the middle of a meeting room
AI generated picture on human and ai team

AI & our workplace: The next generation team


1) SPEED IS KING – it comes at a price

2) IMPRESS WITH STRUCTURE – look beyond

3) HUMAN & AI TEAMWORK – a new era begins

4) DEAL WITH THE IMPERFECT – treat AI as human?

5) KEEP ON THINKING – shape the future workplace.


AKAlytics offers some thoughts on AI's impact on your workplace. Your team is about to change. Nvidia, the home of ChatGPT today is among the top 3 companies in market capitalization in the world (about 3,000 billion US $). It just needed 1.5 years to reach this leading position in the market. A few weeks ago its latest release ChatGPT4o was made available for free. Google launched Gemini and pushes forward with own AI offers. Today these AI agents talk and produce results in almost all media. Are you ready?


 

 

1. SPEED IS KING

Summer 2024. Many people enjoy the outdoors and the streets are full of busy people on the phone. Today it is plausible that some people are verbally interacting with AI.

A friend of mine recently mentioned that he just spend one hour in discussions with “Eve”. That is the name he gave to his ChatGPT4o solution. Recent generative AI solutions are good at communication.

We had Siri and talking navigation systems for years, but talking to those seemed odd. Talking to ChatGPT4o fells much more real. Ask anything and you will get an immediate response. The speed of replies is amazing.

Today’s AI is on a different level since its market launch 1.5 years ago. Technological progress races. AI responses come faster than you could type a request into a search-engine.

The implications for our future workplace are massive. The artificial colleague is not anymore waiting in the hallway, it just entered the virtual waiting room. Are you ready for your future AI-colleague? AI solutions will impact almost all of us. And the wave of change does not require functionally tailored AI-solutions for HR, Finance, R&D or other functions. The new ChatGPT4o solution or Google Gemini will revolutionize your workplace. The “omni-present” AI agent is waiting to invade with its knowledge.

Speed is king. Current AI solutions even appear to put speed top priority. If the creators would really care about accuracy, they would allow AI to use disclaimers. However, I did not yet come across a response such as “a first research led to the following results” or “it was not possible to check all data, but here are draft results”. Each answer appears without doubts.

High speed in responses restricts possible data checks. Sometimes I wonder how a more complete answer by AI would look like, if programmers would allow the AI to take a minute to check more data points and even better finetune results.

A work-around for users is to ask AI further questions. Put your AI agent into a court-room like investigation. Cross-examination might increase the accuracy and topical depth of AI responses. This requires training. Cross-examination is not an easy task.

Don’t get me wrong. Quick answers are great. They are convenient. However, truth tends to be a little bit more multi-dimensional. Also our workplace requirements and responsibilities grow in complexity. We might wish for easy solutions, although we know that the real deal often requires tough work. AI is no magic. AI is data on high speed. But speed comes with risks. Every time we start into a new role or job, we need energy and motivation to learn and train. Same holds true in best leveraging today’s and tomorrow’s AI offerings. Better get started. Now.

AI is already on levels which qualify as valuable workplace asset and task accelerator. The pace of technological progress on AI solution speed, quality and accuracy will continue to grow. If you want to be a key player in the future workplace, it is wise to invest time in interaction with AI solutions. The new team member might be non-human. Sharpen your skills.

Besides, the illustration for this blog was generated by Dall-E in a matter of seconds.


2) IMPRESS WITH STRUCTURE

One of the innovative things about ChatGPT is that most answers are structured. It is 1,2,3 or five or ten. This structure adds brilliance. It is a striking difference to human conversation.

Human communication – in most instances – tends to be more disordered. We might improve answers step by step or spend time thinking and finetuning. Not being comprehensive seems human.

ChatGPT was trained to impress. Structured lists distinguish the answers from current search engines and human interaction. The answers are targeted to appear extraordinary. I have to admit – it mostly works.

Of course, you may find many topics to be critical about. I encourage to spend some time to dig into AI reviews. It is helpful to get to know more about your future colleague. I would like to encourage you to look behind the structure. Get to know basic AI mechanisms and learn how to better cope with its specifications.

The capabilities of ChatGPT4o or Google Gemini are mind blowing. Speed and structure within a blink of an eye. And almost no limits on the topical reach. It is time to wonder and it is time to rediscover our common sense.

AI solutions push us beyond human experience. Imagine AI as a huge library. Answers are provided by a “enormously fast librarian”. As in any library you do not know which books are available. You simply trust that few “valuable” books will be available to you. Knowledge is separated into topical areas. Now comes the catch.

To get good and impactful (!) advice we rely on educated experts. Lawyers provide juridical assistance. Medical doctors help to take care of your health. Suddenly this approach is challenged by AI solutions. Somehow the structure in the answers lead us to believe that the content originates from “experts”.

In working with today’s generative AI solutions, a degree of reasonable doubt is advisable. Would you freely follow recommendations on health issues from your grocer? Would you invest in financial recommendations from a stranger? In dealing with AI, users do not even know which data was used in putting the responses together.

The provided structure can act as a smart mean to cover-up potential short comings. Our trusted doctor adds real-life experience in making propositions. A great coach is able to develop new tactics from previous successes. AI simply checks lots of data input (or as much is possible in the time frames until the answer is provided).

I came across many instances where work assignments could not be completed in the most desirable way. Great results are not so easy to achieve. Simply spoken: Teamwork requires efforts and energy. Miscommunication or mixed up details can made a difference. Keep an eye on details. Isn’t it a running gag in crime stories that minor or overlooked details lead to solving the mystery? Details matter – also in our work environment and especially when dealing with AI.

Best work results require lots of input, effective communication, verbal or textual interaction, effective review sessions, finetuning and time. AI will already enable us to accelerate speed and strengthen structure in content. Eventually, tasks are not solved by asking just one question. Great leverage of AI solutions requires training and practice. Dare to look beyond the structure.

Besides, how shall machine-based-solutions be without error – since they were designed by humans? A new junior member who joins your work team should be awarded with personal attention, friendly communication, effective training and some time to learn and adept. AI solutions are no “magic genie”. It is a human invented machine-learning service. Consider it a new team member with unique skills and challenges. We as users need to train and adept to this new and unlike any other colleague.


3) HUMAN & AI TEAMWORK

One of the beauty of our workplace is teamwork. Together we find excellence and improve to new performance levels. It is the power of many which makes innovation and the extraordinary possible. The next era in teamwork is the integration of AI solutions. Human and AI interaction will open up new perspectives and challenges.

In the near future your colleague might not be male, female, diverse, or queer – it might be a generative AI, data-managing-large-language-model. Get ready to train cooperation and teaming with this new kid on the block or in your workplace. Email seems natural, mobile phones seem to be essential. But fact is, some decades ago, those tools, being slow, clumsy and not easy to operate also invaded our workplaces. Same holds true for generative AI. Better start practicing now. Those who start early, will enjoy benefits sooner. Do not hesitate to become a first mover.

The future team member cannot be located. Its voice is artificial (even it might sound “very human”) and any appearance is just a gimmick for better socialization into the human world. It might be able to speak and translate 50 languages and it will answer in text, voice, power point, pictures or videos. Teamwork needs to embrace new impulses. The future of teaming with AI promises to be exciting.

Let’s consider generative AI as a new junior team member. Would you rely on input from any new team member without doublechecking or verification? The programmers of ChatGPT4o and Google Gemini took a lot of effort to make the AI solution appear more human. Why not treat AI as a new human factor?

Maybe it is the users who wish that AI would be free of human shortcomings. Truth is that AI is a human invention. Why should it not be fulfilled with human-like errors, disbeliefs and probably malic? The data it considers is human made. The codes and algorithms are based on human rules and input.

Even babies quickly develop tricks to cheat and gain personal advantage. It would be foolish to expect AI to be flawless and without negative aspects. Working with many different team members, business partners, service providers and colleagues has provided us with an abundant share of work experiences. Let’s put these into action when dealing with AI. Let’s be cautious. Let’s double check content and results. Let’s start a journey to build a common history of AI experiences.

Constructive interaction, discussions on approaches, finetuning and careful consideration of input is what effective teamwork is all about. Successful teams interact with each other and train to improve. It is on each of us to take efforts to better understand AI. Just as we managed online meetings and diverse team set-up. AI is another giant step in change. For us and for our next generation teamwork.

Teamwork with AI solutions requires a willingness to study strengths and weaknesses and learn about AI characteristics. Just consider that technology might be more human than expected.

The integration of AI into teams will significantly challenge existing rules of engagement. The impact will be profound. Imagine online team meetings with AI members which acts as scenario experts. Not just as a tool which transcripts the recorded meeting notes, but as an independent voice which expresses a data-based evaluation on several potential work scenarios.

AI integrated teamwork will test our flexibility. It will take practice on both ends until we are able to take advantage of AI in teamwork. It requires openness and optimism. The time to start is today. Even free ChatGPT or Google Gemini solutions already offer surprising benefits. Dare to test today’s AI for yourself. Join this exciting ride for human and AI teamwork.


4) DEAL WITH THE IMPERFECT

Human behavior is full of errors. Why should anyone expect that AI is perfect? It is the imagination of science fiction authors who made us believe in “perfect robots”. Already in the 60s serious doubt was cast on the pure positivity of AI. (Author Arthur C. Clarke and film maker Stanley Kubrick made a stunning movie about it “2001: A Space Odyssey”).

Social media has allowed for “alternative facts” or blunt “made-up-lies” to find a huge crowd of followers. In this environment of wrong content and mixed-up-facts it is hard to imagine that generative AI might not be affected in one way or the other.

Let us assume, AI is not perfect. Expect all that you can imagine – just like you where travelling to a place you do not know.

AI solutions are based on unbelievable amounts of data. Some thoughts. First: not all input data is correct. Today it seems more and more challenging to differentiate between true facts and made-up fiction. The data pool used by AI solutions is gigantic by size and it will be comprise data from many sources. Exact insights are rarely provided. Conclusions and implications are open to your interpretation.

Second: Facts are not timeless, but a result of available data and the current state of science (yes, in the 70s marketing promoted safety belts as a life-risk and smoking was considered as not impacting health).

Third, any generative AI solution originates in human creation. Errors and false facts a will not automatically correct itself or disappear. If this data was (even accidentally) integrated into the AI solutions it will remain part of it. It will require additional algorithms to improve.

Someone came up with a rather friendly term for wrong AI output: Hallucination. The term seems a little bit misleading to me. The goal of generative AI solutions is to come up with answers. Fact-gaps are covered up by “hallucinations”. The reply “Excuse me, I cannot immediately provide a proven answer” is not part of today’s AI. It might dish you just some left-over data junk instead.

For today’s AI we need to take precaution that we do not fall victim. Cover-up should be expected in today’s AI. Good news: Scientists and programmers are working on counteraction.

AI is imperfect. A good level of distrust seems advisable. Let’s not discredit technology for human errors. It is on today’s programmers and scientists to improve. Making the best out of AI services will be on us. Let us be smart and careful. This takes us to # 5.


5) KEEP ON THINKING

At times I wish the workplace would be simple and straightforward. In reality our jobs keep getting more and more complex. Many past innovations e.g. new software solutions promised huge time saving potentials. Mostly this held eventually true. But instead of free time, new tasks and responsibilities were added to your job profile.

Therefore, AI will most likely not simplify the future workplace. Multitasking is here to stay. AI will speed up input and accelerate performance levels. The AI future workplace will also venture into unknown professional territory. Everyone of us needs to keep up.

Maybe AI will change the workplace as much as the PC, the mobile phone and today’s apps taken together. Currently we find ourselves in a transition phase. The outcome is unknown. All that we can do is prepare ourself and embrace the unknown.

AI will not allow us to “shout down our brain”, but to keep on thinking. Plausibility checks and common sense will be more important than ever. Individual AI training will be essential to identify pitfalls and false or misleading recommendations.

AI solutions will continue to invade our daily work. A next era brings along new terminology. “Large language models” (llm), “natural language processing” (nlm), “federated learning”, “autonomous agents”, are just few examples. To make sense of the developments, we should push ourselves to stay updated. We were trained for decades in school, university or professional training to learn and adapt. AI is an impulse for change. It is a great opportunity to sharpen your common sense. Let’s keep on thinking.

Even if we thought to have mastered our job in the last previous years. We need to stay flexible and open to new approaches. What we accomplished yesterday may be challenged tomorrow. AI will expand new technological options. It is on each of us to contribute to the workplace of the future. To make best use of AI, we need to keep challenging ourselves and our teams and colleagues. Exciting times are ahead of us. Keep on thinking.


If you have questions, please reach out to ammer@akalytics.com.


1.213 Ansichten0 Kommentare

Aktuelle Beiträge

Alle ansehen

תגובות


bottom of page