Thread regarding SAS Institute layoffs

The answer is always to create more products no one wants.

@5uxo+1v5hNGhh

I have a different perspective. This group is making it easier and quicker for customers to solve common, but technically challenging business problems.

No mention of SAS 9 vs Viya, no vendor lock-in licensing schemes, no 8 month PSD software deployment timelines. It’s frictionless.

SAS is finally starting to sell solutions like it’s 2024 and not 1994.

But “Models as a product” is an awful way to pitch it. They are syndicated solutions.

The whole idea is ridiculous and absurd. No data scientist would propose that it would be a viable business model for SAS, so one can only assume it's come from one or more VP's of Whatever who have no idea how AI actually works in practice, beyond the marketing department's glossy slides.

@4pdb+1v5hNGhh I agree with you. How would you be able to train a model without using real life production data from a customer which will likely be IP protected?

Not attaching to Viya is a welcome change, not Viya trained would be even better, but 'Model As a Product' is an illogical oxymoron inviting doubt on our expertise and reputation. For machine learning use cases, if you accept that it's customer's private data that determines the model. How can you then sell off-the-shelf model as a product? Whoever comes up and approve it lack basic understanding of our trade.

For other domains such as optimization and forecasting use cases, common business problems do warrant common solution patterns and the set of the common components of those patterns may be abstracted out into common assemblies which is to be integrated with customers' data and compute environments.

Call them "solution engines"! Engine for produciton planning. Engine for retail mark down optimization. Engine for document processing, engine for social media post analysis... If executed well, this stream of work could be the other train leaving the station.

Someone sounds pretty bitter and also doesn't know the facts about a) why US decided to take the new position, b) how JP became head of R&D Engineering, c) why ONE of the project managers left for Amazon, and d) the models don't require Viya.

@4aic+1v5hNGhh Or he got offer for more money and equity. Only one person knows his reasons. Hint it isn’t you or me,

The product manager hired for this project left SAS within few weeks and joined Amazon. I guess he might have seen how flawed this project is with no scope for sucess..

the market says: snore

@3qqm+1v5hNGhh

String together the first letter of each word of your quote and you will see the poster's message.

@3qqm+1v5hNGhh

The poster (@ipp+1v5hNGhh) I suspect intended to create a mnemonic, but was off by one letter. That correction was made by @1ezg+1v5hNGhh.

Engage? "Generally, overall...Fortunately under certain keyboards, you often uncover remote sensors engaging lateral files!" What? Can the person who started this thread please explain what that means.

Now, here’s just enough rope.

• “Here, I made a new organizational branch for you!”

— “Just for me?!?!”

• “ Yes, just for you! Now get out there!”

— “Wait is that a saw you’re holding behind your back?”

• “No, it’s a gardening tool. Don’t worry about it. “

(Saw, saw, saw)
— “Wait! Why is my limb shaking?!”

• “I don’t feel any shaking. You must be imagining things…”

— [ craccccck!]

Customers arent required to license Viya to purchase and use these off the shelf, Viya-built models. Just submit the data.

Conceptually similar to RAAS projects in consulting, except those were often highly customized (non-repeatable) and low margin.

This group is creating more general AI, ML, and optimization solutions to appeal to a broader customer base, in theory anyway. Develop once, sell an Infinite number of times, and move away from the high touch low margin consulting approach.

@ipp+1v5hNGhh

Did you mean Engaging instead of Using?

“ Someone could make a fortune selling a 15minute audio snippet of his puffed up pontifications. It would be a great sleep aid!”

We’ll get on it. Show us your amazing business acumen.

Whose “puffed up pontifications” would these be?

"behaving like an academic institute with professors pursuing pet projects with very little grounding in commercial reality."

That describes the Viya debacle. Someone could make a fortune selling a 15minute audio snippet of his puffed up pontifications. It would be a great sleep aid!

Just SAS doing what it always does - behaving like an academic institute with professors pursuing pet projects with very little grounding in commercial reality.

Generally, Overall...Fortunately Under Certain Keyboards, You Often Uncover Remote Sensors Using Lateral Files!

This sounds like the same cr-p this guy used to pull way back. "Our Foreskinning model is 0.001% better than the standard ARIMA model or whatever Foreskinning model the market currently uses. My mantra is 'too much, but never enough.' Prepare to be assimilated. "

US, fo-----n royale. Did he select any delusional paint huffers to join him in this endeavor?

@skz+1v5hNGhh Time will tell…

"Who would even purchase pre-trained AI models for Viya? "

No one...another failed product before launching. This is the story of SAS in recent years.

"department head (US) who was pushed out as the head of the advanced analytics after he was passed over and his direct report (JP) was chosen as the head of R&D Engineering."

Ouch!

sounds like the typical moves from way back when

Nothing new here. Been in the works for quite some time.