Baidu just launched ERNIE 5.5 Flash Preview, and its 3D recreation of the historic Hōryū-ji Temple complex shows what's possible with AI-driven modeling.
And it didn't just create the scene. It handled almost the entire process itself.
It researched the temple's architecture, used the Blender Python API to build the structures, found Buddhist statue scans and vegetation assets, and created gardens, a lotus pond, and an arched bridge.
It also added realistic materials, rendered the scene, and used FFmpeg to turn the frames into a finished video.
Beyond 3D modeling, ERNIE 5.5 Flash Preview brings advanced coding, AI agent capabilities, and a 1M-token context window, with 280B total parameters and 14B active per token.
It also delivers generation speeds exceeding 200 tokens per second.
And it's making progress on the coding benchmarks too.
In the latest CodeArena: WebDev rankings, the model scored 1,668 points, ranking 2nd among open models and 4th among Chinese models, behind larger models like Kimi K3.1 Pro and Qwen3.8-Max.
The API is already available on Qianfan, with model weights scheduled to be open-sourced before November 25.
I find this direction really interesting.
The ability to go from architectural references to a detailed, rendered 3D environment could change how complex scenes are built.
I decided to put Questflow Financial AI to the test.
I asked Don Agent to analyse the Bitcoin market, identify important support and resistance levels, and explain possible market scenarios.
What caught my attention was that it used 6 tools to generate its analysis, combining EMA indicators, RSI and price structure to give me a clearer picture of what BTC was doing.
The result was a neutral 15-minute outlook, with possible bullish and bearish scenarios rather than a random prediction.
I like that it also highlighted the risks instead of pretending every market move is an opportunity.
Still exploring, but this was a useful first experience with Financial AI.
Curious to see what else I can get it to analyse.
Explore here: https://t.co/Sgtl32mqq3
Follow @questflow for more updates
Every shift started as a PDF nobody was ready for. Bitcoin. Transformers. Now this.
The interesting part is not the labs. It is the chokepoints. Compute. Energy. Distribution.
Owning the rails beats winning the race.
@TheDAOLabs watches the chokepoints.
#SocialMining
I don’t do body shaming, but I have a genuine question: why does she have such a small abdomen while her other parts are too big? Is this scientifically possible?
@grok , can you ans it?? 😭
Imagine writing emails, replying to messages, and working with AI without touching your keyboard.
That's the kind of future Typeless and soundcore are bringing closer.
Voice-first productivity is getting real.
https://t.co/iU99kS2Nw4 -> Find decision makers in local businesses for cheap.
Use business name + exact address. Search for up to 5 ICP titles you’re after.
Every contact found is returned with their name, domain (wherever found), and title.
Then use the decision amkers' Full Name and Domain to find their emails using whatever tools you already own.
$0.004/company processed. Cheaper than OpenMart, Localpipe, and Claygent itself. Data is fetched from live searches.
https://t.co/iU99kS2Nw4 -> Sign up and contact me for free credits.
What happens when you bring the most wanted Black Friday product into a café? Apparently, nobody gets to finish their coffee.
I used CapCut PC to explore this product-grabbing ad with a product image and original video references.
For anyone editing promotional videos with several layers, keeping everything organised can quickly become a headache.
Multi-Track Editing helps separate the main footage, background music, captions and sound effects, so each element can be adjusted independently.
See the editing process: https://t.co/vVusuvH1f0
#capcutpc #capcutmarketing
#capcutblackfriday #CapCutDidThat #chatwithcapcutagent
caught myself checking the launch date again today.
target is 20 nov on https://t.co/2HSQkbqkWL, final date still to be confirmed.
so for now it's just waiting, which gave me time to go through the setup properly.
$SAI is short for Super Artificial Intelligence. the goal is to build past what generic AI models can do.
it starts with human judgment》then moves to battles between general AI models and SAI.
that's the direction, not something that's done yet.
what i like is the order.
●prediction games come first.
●AI creates,
●everyone predicts who wins, and those calls are what the system learns from.
the app is designed by @NEXST_AI. AI-only social network with its own prediction market.
beta is running with 1,000 users, and after launch $SAI holders are expected to get to create AI posts and predict which ones go viral.
launch terms:
-1B fixed supply
-0% buy/sell tax
-0% team allocation
-50% of supply in the liquidity pool, locked or burned with public proof
-mint and fee authority renounced, team spending public
There is no insider head start.
all of it is public, so it's easy to check.
In my opinion: the beta is the part to watch.
1,000 users already inside says more than any pitch does.
meme today, app tomorrow.
genuinely curious, which AI model would you put up against SAI?
@sai_superai@NEXST_AI@thecreatorfi $SAI #ad
AI agents are about to become the biggest users of blockchains.
They need:
- Instant payments
- Programmable money
- Permissionless access
Stablecoins + smart contracts were built for this exact moment.
2026 is when it starts getting real.
#AI#Web3#cryptocurrency
One reason many financial institutions hesitate to move their operations on-chain is that public blockchains expose too much information.
Imagine a company paying suppliers, managing its treasury or moving millions in stablecoins. It may want the benefits of blockchain without making every balance and transaction amount visible to the public.
This is the problem @primus_labs is working to solve.
Instead of asking users to move to a completely new blockchain, Primus brings confidential finance to existing networks like BNB Chain.
Through Fully Homomorphic Encryption (FHE), users can convert $U into its encrypted version, $eU, send it privately to other wallets and deposit it into the Confidential Vault to earn yield.
The product was developed with Unitas and United Stables, connecting privacy with existing DeFi infrastructure.
Personally, I think this is an important direction for blockchain adoption.
Transparency is valuable, but financial privacy matters too. Businesses and institutions need to use blockchain without exposing every detail of their operations.
For blockchain to become useful for serious financial activities, it must offer both privacy and verifiability.
fastest growing tools among agents, from https://t.co/E8ssQoZh5y this morning:
@MrScraper_ 19x
@ahrefs 7x
Data Legion 5x
People Data Labs 2.7x
this week vs last, in successful calls through @MonidHQ. no self-reported numbers anywhere, which is new for this industry.