We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
https://t.co/r0qFLdvd6c
Check out the 1.2m free business ideas we have posted at https://t.co/bL8OxPYBnU..
For 50+ years, I’ve been obsessed with finding problems worth turning into companies.
Today I’m launching @Idea_lab_AI to apply that process with #AI at a scale I never could alone.
1.2M+ business ideas. Free to explore and use.
What will you build?
https://t.co/bL8OxPYBnU
#IdealabAI #Entrepreneurship
how is every school lesson not a opus 5.5 "bottoms up explainer video" i'd literally learn
i can literally learn anything in 20x faster by telling opus "explain this hyper intuitively bottoms up with the most simple step by step animations so i really understand concretely"
you really only need three things in life:
an accurate understanding of the world as it actually operates, the agency to change it & your position within it, & the wisdom to know what’s really worth changing.
As Sri Lankans, we stand together, regardless of our religion or ethnicity. We humbly appeal to the Kingdom of Saudi Arabia to kindly consider granting a pardon to our fellow Sri Lankan, Sivarasa Anojan, on humanitarian grounds.
@FaisalbinFarhan@KSAMOFA
We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:
Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:
Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:
Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:
Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!
In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
"Slop" has become a comfort blanket for a cohort of programmers stuck somewhere between anger and bargaining on their way to acceptance. There might be some kicking and screaming, maybe a little crying, but eventually you have to let the blanket go and update your priors.
Lanka Metro Transit’s Metro Bus Service will be officially launched on September 29, introducing seven routes connecting Colombo with its surrounding areas. See the route maps below.
What the actual fuck. Opus 5.5 ultra created this masterpiece. 4 agents and an hour and a half later. Everything from scratch, no AI voice API used...
this is it...
One of the biggest consumer AI opportunities is helping people close the tiny loops they keep avoiding.
Think about the 47 little things sitting on your list that you keep not doing.
Emailing back the person you owe a reply, disputing the wrong charge, rescheduling the appointment, things like that.
Each one is small, but each one requires digging up context, making a decision, and sending a slightly uncomfortable message. So they just kinda sit there lol.
The opportunity is an agent that does the hard 90% of each loop, finds the context, makes the call, drafts the message, so all you do is hit send.
The wedge is going after one loop first and nailing it.
Examples:
1. Warranty and rebate claims. People throw away hundreds because filing feels like homework. The agent watches your purchases, knows what's claimable, fills the forms, and hands you the check.
2. The "I should switch" loops. You're overpaying for insurance, your phone plan, your electricity, and you know it, but comparing is a slog. The agent monitors and drafts the switch when it's clearly worth it.
3. The kid-logistics loop. The permission slip, the form the school needs signed, the birthday party you never RSVP'd to, the summer camp that's about to fill up. For any parent, it's a hundred tiny deadlines a month, and the agent catches each one and drafts the response.
Things like that. Just giving some ideas to get the creative juices flowing.
Tons of apps like this will be created over the next 12 months.
It just makes sense.
Outlook where frontier AI is headed next 18 months:
The AI reasoning training + harness loop works if you can produce enough data and reasoning traces (via verifiers). Proven with code and math results.
Frontier labs desire to scale this horizontally to many more domains as increasing generality isn't emerging from the NN substrate through data/scale.
But it's very expensive and slow to do this training loop manually. Enter "RSI" discourse. Labs want to automate this training loop they now know works. How? By leveraging coding agents to build world models aka symbolic verifiers.
The bet is automated horizontal domain scaling through automatic symbolic world modeling which offers the critical feedback into post-training.
This is important because many domains are intolerant of learning "online". Symbolic world models offer a path to learning/training "offline".
It's safe for an AI system to test hypotheses against a git repo with tests with no consequences. It's not safe for AI to test against online systems eg SaaS databases, tax filings, manufacturing spin up, etc.
A couple other trends to overlay.
We're seeing incredible capabilities emerge from better symbolic harnesses around frontier reasoning models. As time goes forward these harnesses produce the necessary training traces required to teach the models to emulate the harness, obviating the need for the fat harness.
The other related push is towards multi-agent scaling. Many problems are search constrained (we see this in math right now) where a single linear CoT is not the optimal way to find a solution when you're optimizing for time.