Olivia
Online
Gershon
@eliedjenna
Creator building on @X | Passionate about the Apple ecosystem
I help companies deploy, secure & manage Apple devices at scale, so IT stops being a headache
Everywhere
Joined February 2026
152 Following
163 Followers
794 Posts
Pinned Tweet
Businesses operating Apple fleets can bid farewell to their IT headaches.
I’m Gershon, a professional who assists enterprises in deploying, securing, and managing hundreds or even thousands of iPhones, iPads, and Macs at scale.
This ensures that your team is no longer overwhelmed by Apple ID issues, endless support tickets, and the constant question of “who set this up?”
If you’re an IT leader who finds that Mobile Device Management (MDM) feels like a full-time job, you’ve come to the right place.
If this resonates with you, don’t hesitate to drop a 🔥.
Siri is so GOOD 😊
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
@neilcybart I explained it here 👇
https://t.co/p8ynQiP9K7
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
@jackson_lo58573 @MKBHD Funny you put it that way. Apple introduced their 1st on-device and server foundation model in June 2024, the 2nd in 2025 and the current one, which is AFM3.
They’ve had it, and it’s only getting better.
@MKBHD I don’t this analogy goes well with this
@iam_elias1 You should read this …
https://t.co/p8ynQiP9K7
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
@mweinbach https://t.co/p8ynQiP9K7
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
It’s so impressive
https://t.co/rubJWOOG7D
Why People Think Gemini Powers the New Siri (And Why They're Wrong) 😑
Gemini is both Google's AI assistant and the family of multimodal large language models (LLMs) that power it.
Google Gemini is a series of frontier AI models (like Gemini 3.5 Flash for fast, high-performance agentic tasks and coding, Gemini 3.1 Pro for deeper reasoning and creative work).
Google's AI assistant/chatbot is also called Gemini, so this is what most people mean by "Gemini" which is the conversational interface (formerly called Bard).
They share the same name ("Gemini"), which is why it gets confusing. It's similar thing with how "ChatGPT" is both the chatbot and powered by the GPT models.
Now, here's the difference btw the models and the assistant/chatbot:
> models = the actual AI "brains" (the underlying large language models)
> The chatbot = the friendly chat interface/the mobile app that lets everyday users talk to one of the AI 'brains' or the models.
What Apple did when it comes to Apple Intelligence and the new Siri is incredible.
Apple and Google formed an interesting partnership which allow Apple to use Gemini's knowledge for on-device AI, the process is called knowledge distillation.
Most people think Siri and Apple Intelligence is powered by Gemini, no, you're wrong.
Apple has its AI models called Apple Foundation Model, that's right. Apple Foundation Model or AFM powers Apple Intelligence.
During Apple's #WWDC26, the 3rd generation of Apple Foundation Model was announced, AFM 3. This is new AFM 3 powers the latest features of Apple Intelligence and the new Siri.
Based, on this I can say this:
> Gemini (model) powers Gemini (AI chatbot/app)
> ChatGPT (model) powers ChatGPT (AI chatbot/app)
> AFM (model) powers Apple Intelligence (Siri AI app)
Now lemme break it down why some people assumed Gemini powers new Siri or Siri AI. To put it simply, the confusion stems from how AFM3 was trained.
AI models are trained on data, and the same time new models can be trained using existing models as well (the distillation method I mentioned earlier)
@Apple used knowledge distillation from Google's Gemini frontier models to help train and refine their own custom Apple Foundation Models (AFM 3).
They basically used outputs and reasoning traces from Gemini models to refine their AFM via knowledge distillation.
It's worth noting that models and their distilled version could end up being exact copies.
For example:
> DistilBERT is a distilled version of BERT, but smaller and faster).
> GPT-4o mini could be to be distilled from GPT-4o.
> Google's Gemini 3.5 Flash is a faster, lighter variant that could've been optimized from larger Gemini models
However, Apple's own models are not direct copies or deployments of Gemini.
In fact, Apple Foundation Models have their own architectures, optimized specifically for Apple Silicon.
Apple used proprietary data and reinforcement learning to train their AFM, and then refined it using outputs from Gemini frontier models.
#AI #Gemini #AppleIntelligence #Siri #WWDC26 #GoogleDeepMind
The new Siri is so impressive
New Siri and I are spending a lot of time together. And yes, I'm already quite impressed. What do you want to know? What do you want me to test? TELL ME!
Last Seen Users on Sotwe
Most Popular Users
Elon Musk 
@elonmusk
240.9M followers
Barack Obama
@barackobama
119.2M followers
Donald J. Trump
@realdonaldtrump
111.8M followers
Cristiano Ronaldo
@cristiano
111.7M followers
Narendra Modi
@narendramodi
107.1M followers
Rihanna
@rihanna
97.9M followers
NASA
@nasa
92.2M followers
Justin Bieber
@justinbieber
91.1M followers
KATY PERRY
@katyperry
88.2M followers
Taylor Swift
@taylorswift13
82.1M followers
Lady Gaga
@ladygaga
73.6M followers
Virat Kohli
@imvkohli
70.8M followers
Kim Kardashian
@kimkardashian
70.1M followers
YouTube
@youtube
68.7M followers
Bill Gates
@billgates
64.2M followers
Neymar Jr
@neymarjr
63.7M followers
The Ellen Show
@theellenshow
62.4M followers
CNN
@cnn
61.8M followers
Selena Gomez
@selenagomez
61.3M followers
X
@x
60.8M followers