Aktualisiert: 11. Juli
Bemerkung: Diese 1.Ausgabe ist nur auf English geschrieben
It was a great pleasure to host the «It’s about tech event» on Mai 18 at CSEM in the SIP. The third event of this new event series by Swiss made software in close collaboration with BaselTech.org caught some interest. With over 80 attendees, the place was packed. For one obvious reason: the topic was AI.
But those who expected to get the usual hype talk about AI eating up our jobs, telling us before we get sick and knowing who to put in jail, probably didn’t get what they came for. Instead: the crowd was offered a reality check. And this is what we learned from the speakers and panelists:
1) Cool down first, get excited afterwards
DSAI is not the first Tech Bros Hype. And it is not even the first time it was hyped. So the easiest way to stay realistic about it is to call it a SALAMI. So instead of „AI will do XY“ — just say „SALAMI will do XY“ and the whole thing starts to sound reasonable again. Of course it had to be an Italien to come up with this. And by the way, SALAMI stands for: “Systematic Approaches to Learning Algorithms and Machine Inferences.”
2) „DS“ is more important than „AI“
The presented cases showed clearly, feeding a world class learning model with poor data leads to mediocre performance. Letting a good enough learning model learn from excellent data improves the outcome way more significantly. Conclusion: Data quality matters, not (only) quantity. This means if you have a clear purpose for what you want the AI to do and the precise data to train it, then you are in the game — and you don’t need billions, but brains.
3) Tech disrupts Tech and this is good for swiss made software (1)
The big question is: what happens to software developers if you let ChatGPT and Co. spit out code ? In short : you gain time. But you need to be a pro yourself to actually really be able to prompt and control the code. Which is actually great news for Swiss made software and bad news for near- and offshore companies, who did exactly this kind of work before. (Curious to learn more on this topic at our next IAT-Event scheduled for June 20 on „Tomorrow’s skill set in the tech world“)
4) Tech disrupts Tech and this is good for swiss made software (2)
The panelist seemed relaxed about BigTech taking it all, stating there is a lot of activity in the open source space and that the area is just at the beginning (more on this can be read here). They clearly valuate the innovative possibilities above the doom scenario, considering DSAI-technology mainly as a tool that can and should be used pragmatically.
5) If Tech doesn’t disrupt Tech: how bad is this for the rest of the world?
The questions from the audience seemed less relaxed though. What if the big guys suck in all the data and start to colonize us with their non accountable black box AI disrupting not only our working but more so also our social and private life as we know it? The question remained unanswered…
…and the discussion needs to be continued as it did during the networking apéro.
Videos and Presentations:
- Presentation by Christian Walter, swiss made software — Nothing is hallucinated, everything is stolen; Presentation: Video - Presentation by André Bieler, Parashift AG: Case Study — The Do’s and Don’ts in creating a machine learning platform: Video - Presentation by Ursin Brunner, TI&M AG: ChatGPT and Co. — real world projects with Large Language Models: Video - Presentation by Tommaso Bendinelli, CSEM: Some Take aways from implementing AI in industry: Video
- Q&A and panel discussion with speaker, joined by Kevin Lang, brix IT Solutions GmbH and Jan Schulte, Magnolia AG: Video
Read also the blogpost (in german) on swiss made software: here
Register now for our next event: It's about Tech #4: Because talent is not enough - tomorrow’s skill set in the tech world