Generative AI slashes cloud migration hassles, says McKinsey partner
- Reference: 1697023871
- News link: https://www.theregister.co.uk/2023/10/11/generative_ai_cloud_migration/
- Source link:
"This is only starting to scratch the surface. As the large language model (LLM) matures, this timeline to migrate workloads to public cloud will just keep going down – and hopefully the migration will also be efficient," said Srivathsan.
She advised that organizations can use an LLM to understand what a system's infrastructure looks like, decipher its weaknesses and strengths, move the workloads, and then apply AI-based tools to understand if the migration actually worked.
[1]
It can also lean on LLMs for tasks such as producing Architecture Review Board guidelines.
[2]
[3]
The McKinsey partner said although many enterprises are just starting to think about implementing AI, 40 percent of those that have already invested are updating their investment.
Srivathsan called the relationship between generative AI and cloud "symbiotic."
[4]
"We absolutely cannot deny that cloud really is needed to bring generative AI to life. And, similarly, generative AI can really accelerate the migration to public cloud, as well as the unlock from it," she said.
The four biggest use cases Srivathsan sees for generative AI are content generation, customer engagement, creating synthetic data, and coding. The latter role is especially useful when working on legacy code written by coders who are long gone from an organization, or when such code needs translating to a new programming language.
She stressed the need to use public cloud instead of attempting to build in-house models as enterprises don't typically have the necessary access to GPUs. Off-the-shelf models are also cheaper.
[5]More and more LLMs in biz products, but who'll take responsibility for their output?
[6]SoftBank boss Masayoshi Son predicts artificial general intelligence is a decade away
[7]AI startup Lamini bets future on AMD's Instinct GPUs
[8]Intel CTO suggests using AI to port CUDA code to – surprise! – Intel chips
Those working in regulated industries, with large amounts of proprietary data or concerned about IP infringement, can put guardrails in place, Srivathsan said.
She also stressed that LLMs will stay in hyperscale environments for the next five to six years until models mature. While many people feel the need to be closer to their compute power, she advised that very few use cases actually need ultra low latency.
[9]
Unless you are Tesla operating self-driving or perhaps operating in real time on a manufacturing floor, it's just not necessary, she said.
There is also no need to require tailored or massive models.
"Many enterprises think they need a Lamborghini to deliver a pizza, you definitely don't, you probably don't need as complex and as big as a model, you don't need a 65 billion parametric model to generate customer support scripts, for example," said the McKinsey partner.
However, she recommended not skimping on an API gateway between an organization and the outside world in order to surface some of the "real-time alerts" if developers are accessing non-proprietary models or data they shouldn't be touching. ®
Get our [10]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZSbGo7DfWY8dPEcwdxh8WAAAAQQ&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZSbGo7DfWY8dPEcwdxh8WAAAAQQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZSbGo7DfWY8dPEcwdxh8WAAAAQQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZSbGo7DfWY8dPEcwdxh8WAAAAQQ&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2023/09/28/llms_business_risks/
[6] https://www.theregister.com/2023/10/09/softbank_ceo_masayoshi_son_ai_vision/
[7] https://www.theregister.com/2023/09/26/amd_instinct_ai_lamini/
[8] https://www.theregister.com/2023/09/21/intel_ai_cuda/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZSbGo7DfWY8dPEcwdxh8WAAAAQQ&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/
And when your house of cards comes crashing down after the generative AI migrates your database into /dev/null, who will you pay to sort it all out?
McKinsey, of course.
They will charge a princely sum for migrating a database back out of /dev/null, though.
dd if=/dev/random of=/opt/your-actual-database-file.db count=10000000 bs=1024
...there we go - now you've turned a problem of an absent database into a problem of linguistics.
[With due apologies to Messers Adams and Gently]
And of course messrs McKinsey will be charging £300/hour while they wait for that command to complete
For an extra £100/hour they will employ an intern to wiggle your mouse to make the 'migration' go faster. (who they will pay the minimum wage, natch)
Why would anyone trust someone from McKinsey?
Aside from the company being a training ground for crooks*, my partner (an ex-McKinsey person) says the last thing anyone there has is hands-on expertise.
*Rajat Kumar Gupta, Jeffrey Skilling, Puneet Dikshit, Anil Kumar ...
Re: Why would anyone trust someone from McKinsey?
They are my best exemplar for purveyors of fine slideware (though others also have form in borrowing my Elsa-from-Frozen watch and telling me the time).
The use of generative AI
At last, someone has been found whose job really could be replaced by a generative AI. Because that vague, buzzword heavy hype speech certainly sounds like it was written by one.
Have the AI folk actually invented an autogartner?
Re: The use of generative AI
Maybe this is a test by El Reg? Can the readers spot the AI generated articles? If so, I'm calling this one - definitely AI.
However, she recommended not skimping on an API gateway between an organization and the outside world in order to surface some of the "real-time alerts" if developers are accessing non-proprietary models or data they shouldn't be touching.
In no instance of the multiverse does this mean anything to anyone.
"when done correctly"
I do like the caveat that it helps in cases when it is done "correctly", without any indication of how you might check or ensure that at the time (although it might be in the paper, I suppose).
I mean, major government procurements succeed "when done correctly", but that hasn't prevented a whole load of completely ballsed-up projects here in the UK, and, dare I say it, quite possibly elsewhere.
I thought that LLMs actually just did a probabilistic analysis of what word is most likely to 'come next' in the sentence, and did nothing much else, so are not 'intelligent' in any meaningful sense of the word. That McKinsey and Co wants to experiment, sorry, use this 'technology' for their customers (is there an example where McKinsey did it to themselves I wonder) reduces them even lower in my estimation.
On the bright side, I only need one more term to get a full line in my buzzword bingo card.
Re: "when done correctly"
Yes indeed. Those are the magic words that enable the idiot that spouted them to hide behind the bullshit if it doesn't work.
It's the business version of "you're holding it wrong".
Honestly, to publicly state that some absent-minded text-spewing robot that doesn't even understand the significance of what it says improves cloud migration should be a one-way ticket to the padded hotel room.
Such deceit is despicable.
It sounds like a house of cards.