Splice CEO says AI emails are killing conversation

AI-generated emails are eroding the richness of workplace dialogue. When Kakul Srivastava, the CEO of Splice, warned that automated messages are killing conversations, she pointed to a trend that many creators already feel in their inboxes: the convenience of AI‑drafted replies is stripping away the nuance, hesitation, and personal texture that make exchanges meaningful. Srivastava’s background adds weight to her observation. She leads Splice, the sample library that producers use for one‑shots and melodic loops, and she oversaw the acquisition of Spitfire Audio. Before Splice she held executive roles at Flickr, Yahoo, GitHub, and Adobe, and her career has repeatedly sat at the intersection of Silicon Valley tooling and creative communities. The samples that flow from Splice have even appeared in hits like Lisa’s “Money” and Sabrina Carpenter’s “Espresso,” showing how deeply the platform is woven into modern music production.
The problem she highlights isn’t just about efficiency; it’s about what gets lost when a machine writes the first pass of a message. AI tends to favor safe, generic phrasing, smoothing out the quirks that signal intent, humor, or caution. When recipients receive a perfectly polished note, they may assume the sender’s tone is neutral or even disinterested, prompting shorter replies or, worse, a reluctance to engage further. Over time, this can thin out the back‑and‑forth that sparks new ideas, especially in creative fields where a half‑formed thought often needs a tentative nudge to evolve.
So what can be done when AI‑generated email starts to dominate the flow? One practical step is to treat the AI output as a rough outline rather than a final version. After the draft appears, pause and read each sentence aloud, asking whether it captures the exact nuance you intend. If a phrase feels flat, replace it with wording that reflects your personal voice or the specific context of the conversation. Another approach is to set a clear boundary: use AI only for internal reminders or scheduling notes, and reserve human‑crafted messages for anything that involves feedback, negotiation, or creative brainstorming. For those sensitive threads, consider switching to a voice note or a quick video clip; the auditory cues and facial expressions restore the layers that text alone flattens.
Implementing these habits doesn’t require abandoning AI altogether. Instead, it creates a checkpoint where the technology serves as a helper rather than a replacement for the subtleties that keep conversations alive. By consciously reviewing AI‑generated content and limiting its scope to low‑stakes exchanges, teams can preserve the spontaneity and depth that drive genuine collaboration—something Srivastava’s warning reminds us is worth protecting.