Follow @nfk_va
Norfolk (Naw•Fok), VA
Joined February 2011
- Tweets6.6K
- Following3.9K
- Followers921
- Likes1.6K
@OfficialLoganK Friendly flag for ya. I've had several Gemini conversations build interactive visuals unprompted this week but they've all been literally unusable. This example:
Buttons and filter there but showing no data or anything. Same vibes every time.
89
This is the way
A few years ago we would have needed to upgrade to $50,000 a year enterprise plan for Airtable.... a year ago.
Today, we are almost done moving everything to a PostgreSQL database.
A couple devs (global, of course) doing the transfer over a couple of weeks.
tell me again why SaaS has pricing power?
53
Claude Code/cowork usage constraints are going crazy right now. One prompt eating 30%+ of my usage limit is wild. We're gonna need a lot more tokens
1
113
Movie pitch: 6 state legislators and a few staff from opposite parties were working late and end up getting snowed into the state capitol building together. Hijinks and hilarity ensue
59
If I was a betting man I'd say Google puts this as a core part of the Aluminum OS. Next gen OS with deep native AI integrations, works across the phones, watches, desktops, tablets, and whatever is the next hardware form factor. Alphabet wins big with this
Google just quietly dropped an AI that runs on your Mobile and doesn't need the internet.
- 270 million parameters.
- 100% private.
- No servers.
- No cloud.
- No data leaving your device.
It's called FunctionGemma.
Released December 18, 2025.
And it does something wild:
It turns your voice commands into REAL actions on your phone.
No internet required.
No data leaving your device.
No waiting for servers.
Just you and your phone.
That's it.
Let me break down why this matters:
Current AI assistants work like this:
You speak → Words go to the cloud → Server processes → Answer returns
The problem?
→ Slow (internet round-trip)
→ Privacy nightmare (your data travels everywhere)
→ Useless offline (no signal = no help)
FunctionGemma flips this completely.
Everything happens ON your device.
Response time? 0.3 seconds.
Battery drain? 0.75% for 25 conversations.
File size? 288 MB.
That's smaller than most mobile games.
Here's how it actually works:
Step 1: You say "Add John to contacts, number 555-1234"
Step 2: FunctionGemma understands your intent
Step 3: Translates it to code your phone understands
Step 4: Your phone executes it instantly
Step 5: Done. Contact saved. No cloud involved.
The numbers that blew my mind:
• 270M parameters (6,600x smaller than GPT-4)
• 126 tokens per second
• 85% accuracy after fine-tuning
• 550 MB RAM usage
• Works 100% offline
But here's the real genius:
Google calls it the "Traffic Controller" approach.
Simple tasks? → Handled locally (instant + private) Complex tasks? → Routed to cloud AI (when needed)
Best of both worlds.
What can it actually do?
→ "Set alarm for 7 AM" ✓
→ "Turn off living room lights" ✓
→ "Create meeting with Sarah tomorrow" ✓
→ "Navigate to nearest gas station" ✓
→ "Log that I drank 2 glasses of water" ✓
All processed locally. All private. All instant.
The honest limitations:
→ Can't chain multiple steps together (yet) → Struggles with indirect requests → 85% accuracy means 15% errors → Needs fine-tuning for best results
But that 58% → 85% accuracy jump after training?
That's the unlock.
Why should you care?
This isn't about one model.
It's about a fundamental shift:
OLD thinking: Bigger AI = Better AI
NEW thinking: Right-sized AI for the right job
A tiny 270M model trained for YOUR app can outperform a general 7B model.
While using 25x less memory. While running completely offline. While keeping all data private.
The future of AI isn't just in data centers.
It's in your pocket.
And it just got a lot more real.
Want to try it?
→ Download: ollama pull functiongemma
→ Docs: ai.google.dev/gemma/docs/fun… → Model: huggingface.co/google/functi…
PS:) Like, Repost and Bookmark!
If this was useful - Follow for more AI breakdowns
LFG
$goog quietly cooked up a monster release for 2026.
Journo who tried new Google AR Warby Parker glasses sent to devs this week:
"phone-like quality in terms of color, vibrancy, resolution. The seas literally parted..."
New Pixelated pod "Gemini Eyes" ep is a must listen.
34
Add in that TPU chips are already deployed on Pixel phones, soon to be laptops with the Android/Chromebook merger. They are positioned to deliver the best on-device consumer experience too.
This quoted post is unavailable.
30
This is so bullish for pixel phones
⚙️ Google announced Private AI Compute, a cloud platform that runs Gemini on Google’s servers while promising that the user’s data remains visible only to the user, not even to Google.
Because, on-device AI hits limits on reasoning and compute, so heavy requests shift to a sealed cloud environment that aims to keep the privacy guarantees people expect from local processing.
The core of the system pairs Google’s custom TPUs with Titanium Intelligence Enclaves, a hardware isolated area that runs code in a way where the host cannot read user inputs or outputs, similar to trusted execution on other platforms but built into Google’s stack.
The connection between the phone and this enclave uses remote attestation and encryption, which is a proof that the code and hardware are the expected ones before any data is sent, a widely used method in secure enclave systems.
Google frames this as “no access”, meaning the data processed in Private AI Compute should not be inspectable by operators or services outside the enclave boundary, and the claim relies on attestation plus key management tied to the enclave state.
Early user features include better Magic Cue suggestions on Pixel 10 and wider language summaries in Recorder, both tasks that benefit from large context and heavier reasoning across personal data yet retain the privacy promise by confining processing to the enclave.
The phone asks the enclave to run a Gemini prompt, the phone verifies the enclave’s identity, the enclave runs the model on TPUs, and only the user can access the sensitive inputs and outputs, with logs and keys scoped to that sealed context.
The closest analog many will know is Apple’s Private Cloud Compute, and Google’s move mirrors that privacy-first split where simple tasks stay on-device and heavy ones go to an attested cloud enclave.
---
blog .google/technology/ai/google-private-ai-compute/
33
@sundarpichai @tulseedoshi
@joshwoodward Gemini feature request: Add the markup function from pixel screenshot for images generated by Gemini. Lots of folks using the on screen text / notation for editing images but you can't notate directly in the app.
12
It's time @BarstoolGruden
7
1,211
Number four right behind San Francisco, Chicago, and San Jose: Norfolk.
People are struggling. Something has to change.
Yesterday, CoStar published a report on multifamily rent trends for July 2025. There are a couple notable takeaways:
1) San Francisco leads the country in rent growth at 5.3% year-over-year, which is 150 bps above the next highest market. This is largely due to a combination of return-to-office mandates, AI company growth, and the new mayor cleaning up the city.
2) In general, the Northeast and Midwest are seeing continued growth while the Sun Belt is lagging, which correlates directly with the volume of new deliveries in those markets.
Huge for people who hate talking on the phone
New advanced AI features coming to Search:
Gemini 2.5 Pro and Deep Search are rolling out in AI Mode to help with more complex questions and research - available now to Google AI Pro/Ultra subscribers. Plus, a new agentic capability directly in Search: AI-powered calling to local businesses, rolling out to all users in the US.
1
2
113
It's not every day a Virginia political story makes its way (close) to the top of Hacker News. Unfortunately it's bad news
