Mike Sallander

@mikesallander

Entrepreneur, Early Adopter, Technology Buff, Growth-Stage Operator, Product Leader

Charleston, SC

Joined July 2009

2.7K Following

1.1K Followers

1.7K Posts

mikesallander retweeted

signüll

@signulll

12 months ago

was having a convo yesterday about how software used to just mediate communication. like insta, basically a dumb pipe. it renders. you scroll, it shows. or photoshop.. it only did whatever it was designed to do. software was passive infrastructure for human expression. but now software is the conversation. you express ideas to it, through it, & it responds with edits, improvements, sometimes better ideas. then it actually builds them. we’ve crossed into a world where tools are no longer just tools… they’re participants with opinions. this should be a core insight for anyone building anything right now.

366

153

33K

mikesallander retweeted

Aaron Levie

@levie

about 1 year ago

We’re only in the earliest innings of what it means to be an AI-first enterprise. Most companies know that work and business models are going to fundamentally change in a world of AI, and are still figuring their way through this. The past couple years we saw the first major roll-outs of AI chat tools in a corporate context, but this is the first year where AI Agents are a meaningful part of the conversation. 6 months ago AI Agents was a mostly foreign concept to most companies, and today it’s the top AI topic, even though most companies are incredibly early. Here are some early takeaways from these conversations: * Constantly learn and experiment: just as Tobi from Shopify recently laid out in his AI memo, companies should plan to be experimenting where AI can be applied to any workflow. Almost every day there’s a new breakthrough in AI that can lead to new productivity gains when applied to an AI Agent. The only way to discover what those will be for a given task, team, or company, is to constantly experiment and share these across the org. * Get the architecture right: most companies have their data in a structure or organization that does not lend itself to AI Agents being able to operate on that information. Legacy software platforms were not built for an AI-first world, and that will become increasingly painfully obvious as companies trying to deploy AI Agents at scale. Companies will need to ensure they have a modern IT stack. * Get the security right: most enterprises are dealing with years and years of access control sprawl, and security through obscurity, where information is only inaccessible because people can’t find it. This of course no longer works in a world where an AI Agent can operate at a million times the speed of a human to find anything that the human is looking for. * Have optionality: on the one hand, too much choice delivers inefficiency, on the other hand AI is moving so fast that it’s unlikely you want an architecture that fully locks you into one particular approach. In any rapidly changing space, which AI exemplifies, you want the ability to take advantage of any breakthrough innovation that happens wherever it comes from. * Find the right work to accelerate: not all work in an enterprise can be automated or augmented equally. It’s important to know the limits of AI Agents, as well as where the reward isn’t worth the time or cost investment. Finding processes that are either highly repeatable (like reviewing invoices or answering support tickets) or processes where significant intellectual horsepower would deliver real value (like coding or preparing sales strategies) are perfect spots to start. * Don’t always automate the existing process: AI Agents give companies a chance to automate work in completely new ways. There are lots of workflows in the enterprise that have been built up over decades and have become ossified. Deploying AI Agents on those workflows may offer some degree of acceleration, but not any new breakthroughs. Consider the situations where you can use AI to finally change the way the work itself happens. These are just a few common patterns emerging in enterprise conversations. I’ll keep sharing more as enterprises continue to move to being AI-first.

376

407

103K

Mike Sallander

@mikesallander

Who to follow

Last Seen Users on Sotwe

Trends for you

Most Popular Users