** PLEASE SUBSCRIBE TO THE CHANNEL TO SUPPORT ME - THANKS ** More info that I could not fit into the Description section of the video: 2 - Major Cloud Outage Apart from CrowdStrike, here are some of the major public cloud and cloud service outages that occurred in 2024: 1. AT&T: In February, AT&T experienced a significant outage due to an equipment configuration error, causing customers across the entire United States to lose network access for over 12 hours. 2. Verizon: In September, Verizon customers faced service disruptions due to issues with their network. 3. Microsoft: In July, Microsoft had outages affecting Azure and Microsoft 365 services, impacting many users and solution providers - CRN](www.crn.com/news/cloud/2024/the-10-biggest-cloud-outages-of-2024-so-far). 4. Salesforce: In May, Salesforce experienced service failures that affected its users - CRN](www.crn.com/news/cloud/2024/the-10-biggest-cloud-outages-of-2024-so-far). 5. Facebook and Instagram: On March 5, over 11.1 million people reported issues with Facebook and Instagram, making it one of the largest outages of the year. Some irresponsible journalism saw the CrowdStrike incident being blamed on Public Clouds or being a good case for Multicloud adoption, when actually, had enterprises used Public Clouds (for solutions, such as VDI and remote management), these could have mitigated the outage or at least, allowed those same enterprises to recover more quickly than they, potentially, did. As to how a Multicloud strategy can help mitigate/avoid an endpoint device outage, I really do not understand this thinking at all, considering whichever Public Cloud one would use, they all end at the same (single) endpoint device, regardless of which Public Cloud was addressing it. And considering there's only a single/finite number of endpoints at each location, a Multicloud strategy really doesn't help at all with this. Not. One. Bit. www.infoworld.com/article/3477669/the-case-for-multicloud-lessons-from-the-crowdstrike-outage.html www.cio.com/article/3475889/crowdstrike-incident-has-cios-rethinking-their-cloud-strategies.html === 4 - Cloud Winners and Losers Losers: * The UK Post Office and Fujitsu Scandal - Covered extensively in the Feb 2024 episode of CSOP - Also, >unfiltered_ episode * Horizon IT System Scandal - inquiry set up in Sep 2020 - just completed - Another article from the BBC about the "final day of the PO Inquiry" www.bbc.com/news/articles/ckg8d443eg3o * Paula Vennells - Chief Exec of the PO 2012 - 2019 - public hearing - Lawyer: "There has been nothing to show that she acted in bad faith" - Yeah, but she didn't exactly act in good faith, either! - I maintain that NONE of these people exercised even common sense - It surely cannot be normal that you can think that there are that many bad employees in the company - A few cases a year, sure. But 900+? Over a period of literally decades?! - The repetition, the consistency in the alleged wrongdoings cannot be attributed to pure conincidence! * What did she do? She threw her team under the bus! - She claimed that they "didn't tell her" about the scandal - All she had to do was turn on her TV set!! - She told the inquiry: Her senior team failed to inform her of key information but said "she has no desire to point the finger at others". REALLY?!! - This was greeted with "groans" from the public gallery. - And when she said she didn't want to blame others, with "laughter"!! * Inquiry has heard from 298 witnesses, received 780 witness statements and dealt with more than 2.2 million pages of disclosure. * BBC "Months after the Post Office began rolling out Horizon in 1999, the prosecution of sub-postmasters began, and lasted until 2015, resulting in one of the most widespread miscarriages in British justice." * Lawyer for the Sub-Postmasters stated: - "People were ruined, people were bankrupted, people were imprisoned, there were atrocious miscarriages of justice, people died. Whether the board and the executive knew of these injustices from the start is an irrelevant diversion." * UK Gov will pay shortfalls from previous system to Horizon = Capture - 1992 - 1999 - replaced by Horizon === 5 - Tech Updates/Interesting Solutions * Storm brewing around Open Source Initiative (OSI) - Open Source AI Definition thenewstack.io/the-case-against-osis-open-source-ai-definition/ - The article discusses the Open Source Initiative's (OSI) recent attempt to define open source AI, which has sparked significant debate and criticism within the open-source community. The OSI's definition has been endorsed by several organizations, but many prominent figures in the open-source world have voiced their concerns1. - Critics argue that the definition is flawed and doesn't align with the traditional understanding of open source. Key points of contention include: - Lack of Requirement for Open Training Data: Critics believe the definition should mandate open training data for AI models. - Flexibility and Future Understanding: Some argue that the definition leaves too much room for future changes and lacks clarity. - Misapplication of Open Source Principles: There's a belief that applying open source principles to AI is inappropriate, as AI involves more than just source code. - Overall, the article highlights the ongoing tension and differing opinions about how open source should be defined in the context of AI === 6 - GenAI Runaway Successes * OpenAI is the clear outright winner in 2024 - Announcing GPT-4o Multimodal Model in May, a day before Google I/O, and in less than 30 minutes destroying all of Google's 9+ hours of announcements, during the rest of the week! - Announcing o1 (Strawberry) with Chain Of Thought * TED talk recommended by Paul Bevan www.ted.com/talks/jennifer_golbeck_is_ai_progress_stuck?subtitle=en
** PLEASE SUBSCRIBE TO THE CHANNEL TO SUPPORT ME - THANKS **
More info that I could not fit into the Description section of the video:
2 - Major Cloud Outage
Apart from CrowdStrike, here are some of the major public cloud and cloud service outages that occurred in 2024:
1. AT&T: In February, AT&T experienced a significant outage due to an equipment configuration error, causing customers across the entire United States to lose network access for over 12 hours.
2. Verizon: In September, Verizon customers faced service disruptions due to issues with their network.
3. Microsoft: In July, Microsoft had outages affecting Azure and Microsoft 365 services, impacting many users and solution providers - CRN](www.crn.com/news/cloud/2024/the-10-biggest-cloud-outages-of-2024-so-far).
4. Salesforce: In May, Salesforce experienced service failures that affected its users - CRN](www.crn.com/news/cloud/2024/the-10-biggest-cloud-outages-of-2024-so-far).
5. Facebook and Instagram: On March 5, over 11.1 million people reported issues with Facebook and Instagram, making it one of the largest outages of the year.
Some irresponsible journalism saw the CrowdStrike incident being blamed on Public Clouds or being a good case for Multicloud adoption, when actually, had enterprises used Public Clouds (for solutions, such as VDI and remote management), these could have mitigated the outage or at least, allowed those same enterprises to recover more quickly than they, potentially, did. As to how a Multicloud strategy can help mitigate/avoid an endpoint device outage, I really do not understand this thinking at all, considering whichever Public Cloud one would use, they all end at the same (single) endpoint device, regardless of which Public Cloud was addressing it. And considering there's only a single/finite number of endpoints at each location, a Multicloud strategy really doesn't help at all with this. Not. One. Bit.
www.infoworld.com/article/3477669/the-case-for-multicloud-lessons-from-the-crowdstrike-outage.html
www.cio.com/article/3475889/crowdstrike-incident-has-cios-rethinking-their-cloud-strategies.html
===
4 - Cloud Winners and Losers
Losers:
* The UK Post Office and Fujitsu Scandal
- Covered extensively in the Feb 2024 episode of CSOP
- Also, >unfiltered_ episode
* Horizon IT System Scandal - inquiry set up in Sep 2020 - just completed
- Another article from the BBC about the "final day of the PO Inquiry"
www.bbc.com/news/articles/ckg8d443eg3o
* Paula Vennells - Chief Exec of the PO 2012 - 2019 - public hearing
- Lawyer: "There has been nothing to show that she acted in bad faith"
- Yeah, but she didn't exactly act in good faith, either!
- I maintain that NONE of these people exercised even common sense
- It surely cannot be normal that you can think that there are that many bad employees in the company
- A few cases a year, sure. But 900+? Over a period of literally decades?!
- The repetition, the consistency in the alleged wrongdoings cannot be attributed to pure conincidence!
* What did she do? She threw her team under the bus!
- She claimed that they "didn't tell her" about the scandal
- All she had to do was turn on her TV set!!
- She told the inquiry: Her senior team failed to inform her of key information but
said "she has no desire to point the finger at others". REALLY?!!
- This was greeted with "groans" from the public gallery.
- And when she said she didn't want to blame others, with "laughter"!!
* Inquiry has heard from 298 witnesses, received 780 witness statements
and dealt with more than 2.2 million pages of disclosure.
* BBC "Months after the Post Office began rolling out Horizon in 1999, the prosecution of sub-postmasters began,
and lasted until 2015, resulting in one of the most widespread miscarriages in British justice."
* Lawyer for the Sub-Postmasters stated:
- "People were ruined, people were bankrupted, people were imprisoned,
there were atrocious miscarriages of justice, people died.
Whether the board and the executive knew of these injustices from the start is an irrelevant diversion."
* UK Gov will pay shortfalls from previous system to Horizon = Capture
- 1992 - 1999 - replaced by Horizon
===
5 - Tech Updates/Interesting Solutions
* Storm brewing around Open Source Initiative (OSI)
- Open Source AI Definition
thenewstack.io/the-case-against-osis-open-source-ai-definition/
- The article discusses the Open Source Initiative's (OSI) recent attempt to define open source AI, which has sparked significant debate and criticism within the open-source community. The OSI's definition has been endorsed by several organizations, but many prominent figures in the open-source world have voiced their concerns1.
- Critics argue that the definition is flawed and doesn't align with the traditional understanding of open source. Key points of contention include:
- Lack of Requirement for Open Training Data: Critics believe the definition should mandate open training data for AI models.
- Flexibility and Future Understanding: Some argue that the definition leaves too much room for future changes and lacks clarity.
- Misapplication of Open Source Principles: There's a belief that applying open source principles to AI is inappropriate, as AI involves more than just source code.
- Overall, the article highlights the ongoing tension and differing opinions about how open source should be defined in the context of AI
===
6 - GenAI Runaway Successes
* OpenAI is the clear outright winner in 2024
- Announcing GPT-4o Multimodal Model in May, a day before Google I/O, and in less than 30 minutes destroying all of Google's 9+ hours of announcements, during the rest of the week!
- Announcing o1 (Strawberry) with Chain Of Thought
* TED talk recommended by Paul Bevan
www.ted.com/talks/jennifer_golbeck_is_ai_progress_stuck?subtitle=en