Because the yr attracts to a detailed, I discover myself reflecting on what an extremely eventful time it has been—particularly within the realms of AI and Microsoft 365 Copilot. The tempo of evolution in these areas is quicker than ever earlier than, and witnessing how Copilot is reshaping the best way we work has been nothing wanting exhilarating.
This yr, I’ve had the privilege of working carefully with clients on Microsoft 365 Copilot adoption. From teaching and coaching classes to innovating extensions and growing proofs of idea, it’s been a transformative journey. Serving to organizations unlock the potential of AI and combine it into their each day operations has been immensely rewarding.
Being actively concerned with the Microsoft Tech Neighborhood—actually one of the best neighborhood in tech—has been one of many highlights of my yr. I’ve had the chance to talk at occasions throughout Finland, Europe, and the USA, sharing information and connecting with passionate professionals from all around the world. Assembly pals, each outdated and new, inside this vibrant neighborhood has enriched my life, and I’m eagerly anticipating extra of those significant interactions in 2025.Thanks everybody who made this attainable!
Trying forward, I do know that 2025 shall be a yr stuffed with much more developments in Copilot and AI. Clients are more and more recognizing the worth these applied sciences convey. Nevertheless, they’ve additionally realized {that a} return on funding doesn’t occur robotically. True ROI materializes when a bigger variety of individuals perceive how and when to make use of these instruments successfully. This underscores the significance of making related use instances that resonate with customers and improve their workflows.
I’m excited in regards to the upcoming occasions that may kick off the brand new yr. On the finish of January, I’ll be heading to Tallinn, Estonia, for the Cloud Know-how Townhall Tallinn (CTTT). It’s all the time inspiring to begin the yr by participating with fellow tech fans and exploring the newest improvements.
In early Might, I’ll be flying to the USA to talk on the Microsoft 365 Neighborhood Convention. Alongside my good buddy Adam Deltinger, we’ll be internet hosting a workshop titled “Mastering Cross-Organization Collaboration with Microsoft Teams.” Moreover, I’ll be presenting a session on “Administrating and Governing Microsoft Loop.“ These matters are near my coronary heart, and I’m wanting to delve into them with attendees. For those who’re focused on becoming a member of us, you possibly can register for the conference and save $150 with the code SAVE150.
As we method the vacation season, it’s time to wind down and take a well-deserved break. The yr has been stuffed with laborious work, studying, and development, and now it’s vital to relaxation and recharge. I’m grateful for all of the experiences and connections made this yr and am obsessed with what lies forward.
Wishing everybody a peaceable vacation break and a contented new yr. Right here’s to embracing the alternatives that 2025 will convey and persevering with our journey collectively within the ever-evolving world of AI and expertise.
Did that sound like written by AI? Most likely sure and hopefully not an excessive amount of. I intentionally used the Azure OpenAI Service o1-preview mannequin to put in writing this text, based mostly on the data I gave to it, and did little or no adjustments / additions. This reasoning mannequin is simply glorious! I’ve been attempting it out on a number of events and I’m actually impressed on what it could do.
After the primary reply, I requested for a rewrite. Immediate is only a move of my ideas and never by any means a refined one.
GPT-4o is sweet, however o1 is manner higher with its 200K context functionality. Solutions I get from it are longer and higher, it could perceive far more complicated inputs – I seen if I simply write prompts identical to I’d be talking them.. it could give me nice outcomes. It is extremely straightforward to see how this shall be altering the scene throughout 2025, as will probably be increasingly used. The mannequin is at present within the preview however as soon as it goes to GA it’s fascinating to see how will probably be integrated to Microsoft 365 Copilot. GPT-4o is quicker (o1 isn’t sluggish both!) however there’s a clear leap within the high quality.
And it’s acknowledged in Azure AI Foundry fashions catalogue, there are very superior conditions the place o1 will excel
The OpenAI o1 collection fashions are particularly designed to sort out reasoning and problem-solving duties with elevated focus and functionality. These fashions spend extra time processing and understanding the consumer’s request, making them exceptionally robust in areas like science, coding, math and comparable fields. For instance, o1 can be utilized by healthcare researchers to annotate cell sequencing knowledge, by physicists to generate difficult mathematical formulation wanted for quantum optics, and by builders in all fields to construct and execute multi-step workflows.
Utilizing the o1 to put in writing a weblog article is only a small a part of what it’s able to. How you should utilize it then? Similar to another mannequin in Azure AI Foundry – deploy it and use it to energy your customized brokers! In the intervening time you must apply for entry, so you will have a great use case to get began with this.
Pricing is increased, as o1 will use far more computing energy than different fashions. That’s one cause why you shouldn’t all the time use probably the most highly effective mannequin, however as an alternative select the one assembly to your wants. Updated pricing could be discovered at https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/#pricing
Evaluating that to GPT-4o we will see o1 is roughly 6 occasions as costly within the enter and output.
When you may have a great use case for o1,the place outcomes are vital, then most likely the worth isn’t the most important issue whether it is saving individuals’s time massive time – and/or delivering higher and extra right solutions in comparison with inexpensive fashions. Maybe you’ll have 1000 individuals utilizing that a number of occasions a day, and also you begin calculating how excessive the fee shall be. Bear in mind to additionally estimate how a lot it’s going to save time daily for these 1000 staff, and thus rising productiveness.
However the level of the pricing is: don’t simply use o1 as a result of it’s there. Select the mannequin that meets your wants. For instance there are o1-mini, gpt-4o, gpt-4o-mini and so forth. There are , in the meanwhile, 1820 fashions in Azure AI Foundry’s catalogue. For those who want the reasoning and superior mannequin, o1 is definitely one thing you wish to experiment with and check out!
Revealed by
I work, weblog and discuss Future Work : AI, Microsoft 365, Copilot, Microsoft Mesh, Metaverse, and different companies & platforms within the cloud connecting digital and bodily and other people collectively.
I’ve about 30 years of expertise in IT enterprise on a number of industries, domains, and roles.
View all posts by Vesa Nopanen