Technology Labs Tom and Rik

Former member of our Podcast team and Microsoft Cloud Advocate Daniel Laskewitz is now our guest. A great opportunity to check with him the status of Gen-AI at Microsoft. Before we dive into details with Daniel there are some very interesting tech items to discuss. We start with an article on the Verge telling us about how Microsofts Dall-e can filter out specific prompts that could let Copilot generate controversial images (https://www.theverge.com/2024/3/10/24096451/microsoft-blocks-terms-that-could-let-copilot-generate-controversial-images). Local Gen-AI with RAG is something we have been looking at for a long time. Specifically with privacy heavy data (https://blogs.nvidia.com/blog/what-is-retrieval-augmented-generation/). A big topic that Microsoft also implements at scale. Further Gen-AI features that Microsoft delivers can be found in the new partnership that is out there with Mistral AI (https://azure.microsoft.com/en-us/blog/microsoft-and-mistral-ai-announce-new-partnership-to-accelerate-ai-innovation-and-introduce-mistral-large-first-on-azure/). It really unlocks new Large Language Models to be used by developers.

All-in-all a fun episode to listen to, with some great Gen-AI updates from Microsoft.

Our guest is Finn Lorbeer (Director of Engineering at MOIA). With him we find out why one great performing autonomous car can master autonomous driving. This opposed to a fleet of cars that are 'semi'-autonomous. There are some great examples of traffic jams caused by deadlocked autonomous cars (either stand-alone or with each other).

Our podcast is not complete without tech updates and in this episode we discuss the gadget guide of 2023 (which turns out not to be so innovative) and how AI can help Formula 1 to guard 1000's of track violations in one race (a digital referee of sorts).

Gadget gift list EoY 2023 

https://www.theverge.com/23933231/smart-home-tech-gift-guide-2023-video-doorbells-plugs-robot-vacuums 

 

Formula 1 uses AI to observe track limits 

https://www.reuters.com/sports/motor-sports/formula-one-trials-ai-tackle-track-limits-breaches-2023-11-23/#:~:text=The%20AI%20will%20sort%20out,and%20speeding%20up%20the%20response. 

Automate every step in your product creation proces so you can focus on real creativity. Every step along the way of creating a piece of code and eventually an integrated product can be automated. The creation of the (embedded) code itself is the creative part but all the rest: let's automate!

Together with Richard van de Laarschot (chief solution architect quality assurance at Capgemini Engineering) we explore how smart we can make hyper automation. A good message: Be aware not to fall in the bias trap that with a hyper automated production pipeline we find out that the developer working on the most complex parts is responsible for the most defects.

The two tech topics for this episode are:


Extreme water recycling: https://www.wired.com/story/are-you-ready-for-extreme-water-recycling/
European Union Artificial Intelligence Act: https://www.europarl.europa.eu/thinktank/en/document/EPRS_BRI(2021)698792 

A bright future lies ahead of us with current AI developments. In this episode we discuss what's happening with AI in Silicon Valley with Andreas Sjostrom (Leading the applied innovation lab in San Francisco for Capgemini). Generative AI as well as large language models (think ChatGPT) pain a bright future for us. Sceptics would say that ethics shoot all AI developments in the foot but Andreas holds a techno positive attitude and tells us all about possibilities for more inclusion using AI.

We discuss the release of Apple's Vision Pro and the marketing absence of the AI abbreviation during its release. The possibilities for the vision pro are endless. Will it be the next generation computer? The end of PC's and laptops as we know it? Food for discussion.

The use of generative AI is a great development and an awesome companion for us. It is developing at an incredible pace. Just keep up and enjoy the ride.

Can Chat GPT help to test a Web page?

With this question Paul started his exploration of what is possible with current generation AI. His Youtube video shows his experiment and in this podcast we talk about all the possibilities of adding AI to our test toolkit. Just as with test automation it can speed up so much of the test activities we do. Having a 'robot' at your side, helping you out with the obvious stuff gives you so much focus on the creative and volatile aspects of test engineering that is almost a crime not to use it.

Of course there is room in this podcast to talk about tech updates. The James Webb telescope is producing awesome images of the rings of Uranus and there is the imminent launch of JUICE. Keeping in with the space theme we also discuss Star Wars celebrations and the releases of some new Star Wars series and movies!

Again a podcast you do not want to miss out on!

Can ChatGPT help to test a webpage?

Can Chat GPT 4 help us with testing? Could it DO the testing for us?

Star wars celebrations 2023 London 

https://www.theverge.com/23672874/star-wars-celebration-2023-trailers-news 

James Webb telescope pictures of uranus 

https://www.nasa.gov/feature/goddard/2023/nasa-s-webb-scores-another-ringed-world-with-new-image-of-uranus 

The carbon footprint of worldwide IT already represents 1,5x of civil aviation and it is expected to increase in the upcoming years. A good starting point to look for more sustainable IT and reducing that footprint. We talk with Wilson Camargo Junior (Sustainability and Intelligent Industry Consultant) and discuss the options for IT solutions that have a lower impact on for example CO2 emissions. Simple actions as using your phone a year longer can make a difference. And we can take up that action. It is better to mine the heap of discarded electronics in stead of mining our beloved planet.

More info on sustainable IT in this report