AI engineer • Msc Comp Science • Investor Top Rated Plus @Upwork Freelance tips, personal finance, AI
Start investing - FREE stocks
Joined February 2023
- Tweets4.3K
- Following134
- Followers4.4K
- Likes6.8K
I've had so much fun with the new hobby project.
3D model:
The Church of Saint Sava, Belgrade Serbia 🇷🇸
Brings me some bad memories.
I learned the lesson few years ago.
Hardware wallets are still the safest, despite recent hacks.
Crypto security is really risky lately.
First Ledger, as the most popular hardware wallet.
Now Binance, the biggest exchange in the world.
I see why people go with crypto ETFs.
Imagine ETFs custody being hacked.
Worst thing is that I can totally imagine that happening.
Deepseek is highly profitable.
Insane how efficient they are on the tech level.
We all doubt AI valuations, but there are so many unknowns.
Imagine all the Capex returns in a decade.
NFA.
I stay invested.
“A single cluster currently serves 155 trillion tokens/day, generating $38 in revenue per GPU-hour (at off-peak pricing).”
Huge flex from DeepSeek. Not only did they open-source their inference engine, but they also open-sourced their revenue numbers.
They didn't disclose which GPUs they're using here, but DeepInfer supports both NVIDIA Blackwell and Huawei Ascend.
Even assuming they're renting B300s at $7 per GPU-hour, that's ~82% gross margin.
I have to admit I wasn't a huge fan of $SOL few years ago.
The marketing and early investors looked like shitcoin pump and dump.
But engineering behind it is actually very smart.
If your engineering brain hasn't rotten cuz of AI, check it out.
It just realesed a huge update
I own real estate in Serbia.
The market is so hot right now. I didn't realize how bad it is until I saw prices in 🇦🇹Graz, Austria.
Decent area 50m2 apartments:
180k EUR - 🇷🇸 Belgrade
180k EUR - 🇦🇹 Graz
Median salary:
850 EUR - 🇷🇸 Belgrade
2500 EUR - 🇦🇹Graz
For anyone who wants a free Claude Code alternative:
opencode.ai/zen
Plenty of free models to choose from.
And it's been like this for months. Not some temporary promotion.
Should be enough for regular daily work.
For engineers interested in always on agents.
Stanford course:
youtube.com/watch?v=L1OBRtfP…
Regular DCA executed.
Bought another ~$30/mo passive forever.
Doesn't sound like much, but it adds up and compounds if reinvested.
Let's think rationally.
The cryptography penetration risk is real. On the other hand, AI might break it but also fix it.
So it's either eventually going to $1,000,000 or to 0.
Where to focus If I'm starting my journey now.
I'd add unique expirences here tho.
I'd expand the moats left to:
- Proprietary data or content: data nobody else has, or unique content you can create
- Network effects / community: users make it more valuable for each other
- Audience / distribution: you can get attention on demand
- Physical component: hardware, locations, real-world operations
- Capital: owning assets and investing (esp into the AI companies as they are and will suck up large parts of the economy)
- Regulatory: licenses and legal barriers
Insider projections from frontier AI labs:
The industry value is going up by 30% per year over the next 10 years.
If you are afraid of being replaced by AI - stay invested as a hedge.
Ignore AI bubble speculations.
Btw if true, that would be the largest industry move ever.