https://youtu.be/dZxyeYBxPBA?si=09AvHbklIct55iWZ
The level of AI impact ON SOFTWARE DEVELOPMENT continues to be profound. What is the probability that even the most leading edge (something SAS has not been for a long time) traditional software vendor can survive for even 5 more years under this level of automated intelligence -/ WITHOUT RADICALLY changing their business model and approach?
"Banks, pharma, and health care are going to be highly cautious"
Many unregulated customers will be cautious too. People don't like change, and businesses don't like risk.
In the long term, traditional vendors may well be doomed by AI. In SAS's case, the retirement of elderly SAS users is probably a bigger factor.
In the short term, I bet the worst impact of AI is to give customers another reason to switch from SAS. They already had Open Source; they already had the Cloud; the hyperscalers' AI platforms gives them another reason.
If enough customers switch, to maintain its profit margin SAS may need to reduce headcount by ~4-5% each year, instead of ~2-3% as they've been doing.
That's the worst I expect. Not sudden death, but an ongoing decline accelerated by AI.
The best hope for SAS is honestly to position itself as the safe play in highly regulatory environments. Banks, pharma, and health care are going to be highly cautious about letting Agentic AI run wild.
Outside of that, enterprise software in general in doomed across the board as long as you are dealing with publicly accessible data.
It’s all hokum!
Thanks for the link. Jones gives a good summary of the changes that have accelerated in just the past couple of months. If ever there were a moment when startups could disrupt traditional software vendors, this is it.
By definition, traditional vendors have had years in which to hire a wide distribution of talent. As Jones says, these powerful tools make the most talented people more productive.
But less talented people will not be able to use these tools. They'll never learn to specify tasks correctly, or manage multiple AIs, or catch mistakes the AIs make. If you give them these tools, they'll just make a mess.
Over the next several years, it will become obvious who can handle these tools and who can't. And every traditional software vendor will realize that some large fraction of their R&D staff adds no value.
that's an interesting video that seems to be mainly about individual productivity (you are now the dev manager of as many agents as you can orchestrate).
adding that up to SaaS and the Anthropic announcements, we have the "SaaSpocalypse" of the past couple of days (software stocks diving). SAS was disrupted by SaaS and more. SaaS is being disrupted by AI. what happens next?
76.4%