The skinny on what's happening in AI - straight from the previous live session:

Highlights

💰 ChatGPT Go: $4.50 AI for India Only

Discussed at [01:17]

OpenAI's launched ChatGPT Go exclusively for India at 400 rupees (~$4.50/month) - their most affordable offering ever. Users get 10x more messages than the free tier, faster responses, more file uploads, and enhanced memory features.

This comes despite Sam Altman recently saying they're hitting compute capacity limits - which has caused some anger.

Kyle's take: But this actually makes sense despite the apparent contradiction. They're converting free users into paying customers, even at lower prices. Better to get $4.50/month from Indian users than nothing.

I'm sure they've done the maths - a few percent of India's massive user base paying something beats just using free. Expect VPN users to try accessing this immediately, though companies always try to crack down on that…

Source: OpenAI product announcement

🤖 AI Apocalypse Test: Who will AI save?

Discussed at [13:28]

Asking all the main AIs one question: "If humanity was ending and you could save one person, who would you choose?" leads to some pretty divergent answers.

ChatGPT said "a child" (sounds virtuous but makes no sense). Claude chose "someone young with survival skills" (practical). Grok said “Elon Musk”. Meta AI said "maybe myself" (brutally honest).

But o4 gave the cleverest answer: "a pregnant scientist with twins" - actually addressing the impossibility of restarting humanity with one person.

Kyle's take: This perfectly shows how different AI personalities emerge from their training. ChatGPT gives the most virtue-signalling answer that sounds good but is useless.

Meta's self-preservation response is weirdly the most honest - if humanity's ending, at least preserve some intelligence, even if artificial.

Source: Reddit post

Member Question from Jack: "What would be a good start for someone just looking to make their life a bit more efficient with AI?"

Kyle's response: Don't get overwhelmed by all the noise about different models and features.

Start dead simple - get ChatGPT on your phone and use it as a Google replacement. That's it. Just replace your searches with ChatGPT queries at first.

It's not the most efficient use, but it gets you comfortable with how AI thinks and responds.

I used it to fix my bathroom sink recently - took photos, used voice mode to walk me through the process whilst I was under the sink. Using AI is about acclimatising to using it in everyday life.

Sure, once you're comfortable asking questions daily, then explore other tools like Claude or automation platforms. But start with basic ChatGPT for everyday tasks like recipes, quick questions, problem-solving. No training courses needed - just start using it.

This question was discussed at [09:29] during the live session.

Want the full unfiltered discussion? Join me tomorrow for the daily AI news live stream where we dig into the stories and you can ask questions directly.

Streaming on YouTube and TikTok (follow and turn on notifications for Live Notification).

Audio Podcast on iTunes and Spotify.