I love MSFS avid fan but personally I cannot see any improvement in the sim with atc, its never happened always we had to rely on 3rd parties. “Cant say anything right now” in my thoughts nothing planned.
I love my 3rd party ATC, not always american accents transitional alttitude Fl180..I hope I am wrong.
Sayintentions ai having a blast. Its also increased my confidence in speaking to ATC.
Well maybe, but he does go on to say, “but we will in the future.” I’m just going to call it, they’ve either partnered with BATC or SI… or there’s another unknown player who has been quietly working on something.
Or of course, @JWATTS0275 is correct. 
As a RL GA Pilot, SayIntentions.AI is an amazing improvement in MSFS ATC and approaches close to realistic RW ATC. I LOVE IT
Not only that, but its development speed and response to the MSFS ( and X-plane) community, sets a very high bar for other Sim developers.
For me, SI has done more to transform MSFS from a game to a Simulator, than any one other single addon. - and there were already some pretty amazing addons out there, ie LNM
BATC is also still in beta and offer a PAYG model so I don’t see your point
That’s okay. No need for an endless back-and-forth.
probably best
Working Title had ATC down as one of their long-term projects, but they noted that it could only be adequately addressed once certain core parts of the sim were changed. A lot of the fixes of avionics and the development of the avionics framework were essential precursors to better ATC.
My interpretation was that this enhancement was brought across to be part of 2024 development. Dont know if WT still have the reins, but another core stepping-stone on the route to better ATC was better native flight planning tools and we know that they are bringing that to MS24.
yep, and they showed the flight planner, and its not November yet. Im sure they will announce it. What? they only were given an hour but went over that at FSExpo? Im sure we’ll hear something.
I don’t think that SI has much to worry about with WT as competition.
I cannot see MS absorbing the cost of AI, and a non AI solution would end up just being a better version of what we already have in MSFS, or what BATC offers.
Also, forget about the type of day-by-day support that we get from Brian and his Team, if someone else , partnering with MS, is developing an ATC solution.
With BATC and SI already establish in the MSFS market, it would however be interesting (/ amusing) to see what WT thinks they can come up with that would we anywhere a good… the only good feature being that if it is included in MSFS 2024, it may be FREE … but you get what you pay for - something that is clearly very true with a realistic ATC system.
Also, not EVERYONE flying around in MSFS even wants or cares about ATC, as much as they want other FREE stuff, and a sim that does not slow down and crash on them while they are flying.
Not really commenting on the quality of what the ATC will be in MS24, but just that WT gave a development roadmap and revision of the native ATC was on it. They have been sticking to it although they have gone silent on the subject of ATC as of late.
I wouldn’t be too quick to assume MS wouldn’t absorb the cost of AI. For me one of the most significant things Jorg said at the expo presentation was that: ‘we access the MS techstack, and we get it for free’.
There are business advantages to MS in using a platform like MSFS to develop technology which could easily offset any costs (the tech used to develop truly effective AI ATC would have very wide RL application)
Microsoft doesn’t own ChatGPT nor the company OpenAI which is founded by Chatgpt’s CEO Sam Altman. but the both companies have been partnered commercially since 2016 so Microsoft continues to be the company’s largest investor.
That’s a year old post from Who Owns ChatGPT and Its Creator, OpenAI?
Does Microsoft Own ChatGPT?
Microsoft doesn’t own ChatGPT, nor the chatbot’s founding company OpenAI. However, the two companies have been partnered commercially since 2016, and Microsoft continues to be the company’s largest investor.
Microsoft invested $1 billion into the research lab in 2019, after the companies first made their partnership exclusive. Then, after the resounding successes of ChatGPT and OpenAIs’ other products, Microsoft invested a further $10 billion in the company this January, as they entered the third phase of their partnership.
OpenAI’s motivations are pretty obvious, but Microsoft also benefits heavily from this relationship. By investing in OpenAI, Microsoft is able to gain access to the lab’s cutting-edge technology, which it can then deploy across its broad range of products and services.
These heavy investments are paying off too. According to Business Insider, Microsoft’s Azure’s revenue is expected to jump 27% next quarter, compared to the last – with 1% of this growth being attributed to AI services.
It’s a little more complex than that too, as the OpenAI/MS agreements are secret.
ref: Here's why OpenAI clarified that Microsoft is not an owner | Windows Central [from dec 2023, 6 months old]
Microsoft also clarified the situation in a statement earlier this month. “While details of our agreement remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions,” explained the tech giant earlier this month.
…
While Microsoft and OpenAI are intertwined and work together in various areas, the two companies also compete with each other. Microsoft’s massive investment in OpenAI sees Azure as the exclusive cloud partner of OpenAI. That partnership also has OpenAI share IP with Microsoft exclusively. But any form of AGI (artificial general intelligence) that OpenAI creates at any point in the future will not be shared with Microsoft.
MS is also developing it’s own AI to compete with ChatGPT++, so it’s not their only option in the AI solution race. ChatGPT isn’t el supremo either. Anthropic AI’s https://www.anthropic.com/ various Claude models beat it hands down in a number of areas, as one example. This area continues to evolve. It’s also not the solution to AGI, that’s a given.
I wish SI the best of luck, but I’m 99.999999% certain they will continue to struggle trying to prompt wrangle ChatGPT into giving 100% accurate output. It’s a useful technology for certain contexts and use cases that can handle this (eg: very narrow tasks, creative tasks etc), but if you want reliable trustworthy output it’s well nigh impossible at the moment, and may never be solved because it’s a feature of the LLM tech, not a bug per se. We’ll see!
Well, it numerous things to make an AI ATC for MSFS , Including an AI engine, and the AI developer.
maybe Asobo already has an AI developer, the number of positons they are currently advertising for has greatly reduced since the start of the year.
I find it interesting they are looking for both “ATC and Flight Planning” devs AND “Autopilot” Devs.
I thought that WT has both those covered ?
I find it interesting that for the ATC position, the goal is The evolution of the air traffic AI system, but the position requirements do not mention any AI experience.
It’s entirely possible WTT’s responsibilities lie purely with the underlying tech that ATC relies on, of which the latest planning system that was announced is a key part, and a big enough job in itself. They’ve always maintained this was the plan for “fixing” ATC, and they appear to be delivering on it (though Jorg had a different explanation re: the reason for charts inclusion in the base sim, but I think there’s overlap).
There’s no reason why they (MS/Asobo) can’t spin up a separate team(s) to handle any AI on top of that.
And AI can’t work at all if the underlying systems are not solid.
That would be the very definition of GIGO, or in ChatGPT’s case TurboGIGO!
or maybe GaiGO?
It’s interesting watching it develop.
Don’t forget the current early adopter user base has a much larger tolerance for current LLM’s accuracy foibles than a more mainstream user base would have.
You could posit this might not be an issue considering the state of the current base ATC and its well known “issues”, but who knows. We’ll see!
Edit:
Not to mention, AI ATC is a hard problem
Yes, with great anticipation, but not over expectations
I’m trying to join the discord but the link just opens the discord app and does nothing… Could use help or advice?
Hi there, Does SayIntentions have a learning curve?
SI is an ATC service. It doesn’t teach you communication procedures.
If you are already familiar with ATC communications it will be a breeze. If not there are books and online resources that will teach you how.
It does have a “student mode” that slows things down a bit if you are just learning, but you are expected to know the basics before you begin using the service.
So I tried SayIntentions a few weeks ago and was immediately hooked, especially in VR at the airports I learned to fly at.
But I’ve also now come across PSXT + Real Traffic for flying with live aircraft with only a 30 second delay.
Now if only we could use our local controller’s voice offline on our own systems to train the AI speech it’d be perfect real world immersion.
Or even better, if SI could listen in to LiveATC channels and be aware of what is happening in the real world listening in to their comms and then acting as the proxy ATC for you to sequence you. That may be several layers too deep as it’d mean synced data between PSXT/RealTraffic, LiveATC, FlightRadar24 and SayIntentions but boy would that day where that level of integration could happen would be amazing.
I’m a bit surprised there isn’t more discussion on this topic!
Few months ago I tried the ‘free sample’, and then signed up for a monthly subscription - a bit concerned about the price - but easy to cancel. Now, it’s a ‘can’t fly without’ add on for me.
I fly VFR and still a lot to be worked on, but adds a dimension we’ve never had before. Surprising success is the use of live and recorded other users comms. No idea how much filtering they have to do - but I’ve been suprised how good that is … despite the occasional snatch of background music… 
