The best way to think about this AI speech companion is through the child’s comfort, the family’s real routine, and communication support that does not become pressure to perform. Home practice works best when it stays respectful and doable.
Let me tell you about a Tuesday night in February. My friend Carla, whose son Mateo is three and mostly nonspeaking, texted me a video. In it, Mateo is sitting at the kitchen table with a laminated board of picture symbols. He taps “more,” then “banana.” Carla’s mother-in-law is standing behind him, arms crossed, saying (loudly enough for the phone mic to catch it), “He can talk if you just make him try harder.” Carla captioned the video: “This is what I’m up against every single day.”
That video is the actual problem most families face with AAC. Not the device. Not the app. Not even the learning curve. It’s the other adults.
Before anything else: AAC does not delay speech. Schlosser & Wendt (2008) reviewed twenty-three single-subject studies and found neutral-to-positive effects on spoken language development. If an SLP recommends AAC, it is a tool added to your child’s toolkit, not a verdict on their voice. Hold onto that sentence. You’ll need it at Thanksgiving dinner.
The Real Barrier Is Buy-In, Not Technology
Most AAC guidance focuses on features, apps, device comparisons. That stuff matters, but it’s a second-order problem. The first-order problem is: will every adult who spends meaningful time with your child actually use the system? Spouse, grandparent, daycare teacher, weekend babysitter, the aunt who watches him on Thursdays.
If the answer is “only me, the parent who researched this at 1 a.m.,” then the AAC system is functionally a single-room tool. And single-room tools produce single-room results.
The research backs this up. Aided language input from adults (meaning you model on the device, constantly, not just when your child seems stuck) is one of the strongest predictors of AAC success. Romski, Sevcik, and colleagues at Georgia State ran a randomized trial in 2010 comparing augmented input, augmented input plus output, and spoken language only in toddlers. No evidence that AAC delayed speech. In many participants, spoken language increased. But the gains tracked with how much modeling adults did, not how fancy the system was.
Your child’s $8,000 speech-generating device is a paperweight if Grandma won’t touch it.
What Modeling Actually Looks Like (and Why It Feels Weird at First)
A two-year-old hands you a paper card with a picture of milk. Then he points to a card with crackers. Six months ago he could not tell you he was hungry without escalating to tears. That paper system is AAC. So is a free communication app on a tablet, and so is a dedicated device with eye-tracking. The form factor matters less than what the adults around the child do with it.
Here’s the ratio that SLPs typically recommend: model on the system at least ten times for every one time you expect the child to use it. That means you, the adult, are tapping symbols and narrating your own actions throughout the day. “I want coffee.” “Let’s go outside.” “That dog is loud.” It feels performative at first, maybe even silly. You’re essentially narrating your life through picture symbols to a toddler who may or may not be watching.
Do it anyway. Communication is caught more than taught. Your child is watching even when you think they’re not.
The tricky part is getting every other adult to do the same. Here’s what I’ve seen work, practically:
- Start with one person at a time. Don’t hold a family summit. Pick the adult who’s most willing (or least resistant) and spend a week getting them comfortable. Then add the next person.
- Make it physical. Tape a low-tech communication board to the refrigerator, the back of a dining chair, the car seat headrest. If the system is visible, people use it. If it lives in a backpack, it stays in the backpack.
- Give skeptics a single job. Grandma doesn’t need to learn the whole system. She needs to model “more” and “all done” at snack time. That’s it. Two symbols. Twice a day. Expand later.
- Send the research, but don’t lecture. A one-page handout with the Schlosser & Wendt (2008) finding and the ASHA 2021 position statement on AAC goes further than a 40-minute kitchen debate. Some people need to see it in print before they’ll let go of the “just make him talk” instinct.
The Mistakes That Eat Months
These aren’t moral failures. They’re patterns that show up in family after family, and naming them saves time.
Treating AAC as last resort. Some insurance companies and even some clinicians still use a “must fail spoken language first” framework before approving AAC. The literature is firmly against this. ASHA’s 2021 position statement and most current insurance criteria have moved away from that older framing. If someone is telling you your child “isn’t ready” for AAC, push back. The evidence is on your side.
Quizzing instead of communicating. “What’s this? Point to the apple. Show me dog.” That’s testing, not conversation. Your child can tell the difference. Model, narrate, respond. Don’t interrogate.
Modeling only during meltdowns. If the only time you pull out the communication board is when your child is already dysregulated, the board becomes associated with crisis. Model when things are calm, boring, happy. Especially boring. Boring moments are where real communication habits form.
Assuming AAC means giving up on speech. I’ll say it again because the fear is persistent: AAC does not replace speech. It provides a bridge while spoken language develops at whatever pace it develops. For some children that bridge is temporary. For others it becomes a permanent, valued part of how they communicate. Both outcomes are fine.
A Practical Three-Week Plan (Pick Two, Not Six)
If you want a checklist, here it is. But the assignment is: pick two of these. Run them for three weeks. Then come back and add two more. Parents who try to launch all six in week one typically abandon everything by week two. Two steps is the right size.
- If an SLP has recommended AAC, schedule the evaluation. Don’t table it.
- Start modeling on a low-tech option (paper cards, a free app like CoughDrop’s trial) while the formal device process moves forward.
- Model ten times for every one expectation of your child using the system.
- Loop in one additional adult this week. Just one.
- Track what your child requests, comments on, and protests using AAC. These are language samples, not just checkboxes.
- Read a plain-language summary of Schlosser & Wendt (2008) before deciding AAC will “delay” speech.
A note on bad days: five minutes of a routine on a hard day still counts. Skipping it entirely does not. Build a low-effort fallback version so that even when you’re running on three hours of sleep, something happens. Consistency over intensity, every time.
Where LittleWords Fits (and Where It Doesn’t)
I want to be direct about this. LittleWords is a speech-practice app. It is not an AAC device. If your child has been recommended for AAC, please pursue the AAC evaluation. That comes first.
What LittleWords can do is sit alongside a paper or digital AAC system as a low-pressure practice window during the day. Think of it like the difference between physical therapy and going for a walk. The walk isn’t the therapy. But the walk still matters, and it’s better than sitting on the couch. You can read more about the approach and the founder story at this AI speech companion.
A few specifics: LittleWords is currently in a waitlist phase, with iOS and Android launches planned for Spring 2026. Founding Family pricing is a one-time $49 for lifetime access. The app is COPPA-compliant (no kid data sold, parental consent required, zero advertising). It’s designed in collaboration with licensed SLPs, and public clinical reviewer attribution will follow once final credentialing is complete.
When to Call an SLP
If your child is over two with limited spoken language and high frustration during communication moments, request an AAC evaluation. An SLP with AAC expertise will assess motor access, symbol understanding, and family modeling capacity, then recommend a system based on what they find.
If you don’t yet have an SLP, the fastest paths in:
- Pediatrician referral for insurance-covered evaluation
- Your state’s Early Intervention program (if your child is under three)
- Your school district’s evaluation team (if your child is three or older)
- Telehealth speech-therapy clinics, which often have shorter waits than brick-and-mortar
Frequently Asked Questions
Q: Will AAC delay my child’s speech? A: No. Schlosser & Wendt (2008) and multiple subsequent reviews show neutral-to-positive effects of AAC on natural speech development. The fear is understandable; the evidence is clear.
Q: Is AAC only for non-speaking children? A: No. Many minimally speaking, gestalt-processing, and intermittently speaking children benefit from AAC alongside spoken language.
Q: What does AAC cost? A: Low-tech AAC (paper cards) is essentially free. Free apps exist. Dedicated devices are often covered by insurance or schools when an SLP prescribes them.
Q: Should I model on the device myself? A: Yes. Extensively. Aided language input from adults is one of the most important predictors of AAC success.
Q: Is LittleWords an AAC device? A: No. LittleWords is a speech-practice companion designed to complement therapy, not substitute for a clinician-prescribed augmentative and alternative communication system.
Q: How do I get an AAC evaluation? A: Ask your SLP for a referral, or contact a local AAC specialist clinic directly. Many hospital systems and university clinics offer dedicated AAC evaluations.
Q: What if my family members refuse to use AAC? A: Start with one willing adult. Build from there. Share the research. And if someone tells you AAC will “make him lazy about talking,” hand them the Schlosser & Wendt meta-analysis and walk away.
Show up small. Show up often. That is the whole job.
