Serial builder: Web3, AI-to-video, tokenization, RWA protocols
Learn more 馃憠
Joined June 2022
- Tweets11.5K
- Following49
- Followers4.7K
- Likes7.3K
Jan Vidic retweeted
Partner Highlight: @noctocode
Noctocode is a product development partner building across web, AI, and blockchain applications, helping turn early-stage ideas into scalable products.
They鈥檙e part of the build layer of the RWA stack.
Jan Vidic retweeted
GPT-5 and Claude 4.5 are already close to semi-autonomous ML research
when you use them for ML coding, it's clear they were trained more on ML code than on other code
with some creative setup, you could make them nearly autonomous today
but the issue is depth.
they don't explore the solution space well enough, so picking the right path still needs humans
Jan Vidic retweeted
When the CEOs of Nvidia, Samsung, and Hyundai meet and have a shot, you know the future is being negotiated
Jan Vidic retweeted
SciFi fans may have noticed that the @extropic chip resembles the "cortical stack" implants from Altered Carbon.
In Altered Carbon, the stack implants store a person鈥檚 consciousness, memories, and personality.
I'm sure it's not a coincidence.
Google AI Studio Launches Logs & Datasets for Smarter AI Monitoring
馃毃Google AI Studio just launched Logs & Datasets , a new feature that helps developers monitor, debug, and improve AI app performance.
You can now:
- Enable logging with one click (no new code)
- Track API calls, inputs, outputs, and errors
- Export logs as datasets for evaluation or prompt tuning
- Use datasets for testing, performance tracking, and feedback
Note : You need to set up billing to enable Gemini API logging.
Jan Vidic retweeted
馃毃Google AI Studio just launched Logs & Datasets , a new feature that helps developers monitor, debug, and improve AI app performance.
You can now:
- Enable logging with one click (no new code)
- Track API calls, inputs, outputs, and errors
- Export logs as datasets for evaluation or prompt tuning
- Use datasets for testing, performance tracking, and feedback
Note : You need to set up billing to enable Gemini API logging.
Jan Vidic retweeted
Microsoft鈥檚 latest SEC filing quietly exposed that OpenAI lost around $11.5 billion last quarter, based on Microsoft鈥檚 27% ownership stake and a $3.1 billion hit to its own net income.
The filings confirm that Microsoft has funded $11.6B of its $13B commitment to OpenAI, with losses directly affecting Microsoft鈥檚 earnings under equity accounting rules.
Jan Vidic retweeted
Microsoft seemingly just revealed that @OpenAI lost $11.5B last quarter.
Microsoft reported earnings for the quarter ended Sept. 30 on Wednesday after market close and buried in its financial filings were a couple of passages suggesting that OpenAI suffered a net loss of $11.5 billion during the quarter.
Jan Vidic retweeted
馃毃 OpenAI introduces Aardvark, an agentic security researcher powered by GPT 5.
Aardvark acts like an autonomous AI security expert that scans codebases, finds vulnerabilities, validates them in a sandbox, and suggests fixes using Codex.
Key features
- Continuously monitors commits and detects vulnerabilities
- Confirms exploitability in isolated environments
- Suggests and attaches secure patches for review
- Integrates directly with GitHub and existing workflows
Currently in private beta, Aardvark has already identified 92 percent of known vulnerabilities in tests and discovered multiple new CVEs in open source projects.
A new step toward AI driven, defender first cybersecurity.
Jan Vidic retweeted
OpenAI has set a 2028 goal to build a fully automated AI researcher
if they achieved it, this aligns with ex-OpenAI Leopold Aschenbrenner (situational awareness) forecast:
> AI progress won't stop at the human level
> hundreds of millions of AGIs could automate AI research
> decade of algorithmic progress (5+ orders of magnitude) into a year or less
> humanity could move quickly from human-level to superhuman AI
Tiny Artificial Neuron Could Power Brain Scale AI With Millions X Efficiency
Are you ready for one more computing breakthrough? - the 5th in the last 24 hours.
Breakthrough artificial neuron mimics the human brain, tiny memristor design could revolutionize AI efficiency.
Scientists have created a spiking artificial neuron powered by a diffusive memristor, a nanoscale component that behaves like a real biological neuron.
This revolutionary 1M1T1R (one transistor and one resistor) design packs all core neural functions - learning, firing, adapting, forgetting into a footprint the size of a single transistor.
Operating at picojoule energy levels, it鈥檚 MILLIONS of times more efficient than current AI chips. 馃憖
This breakthrough could enable neuromorphic chips with billions of brain like neurons, capable of real time learning, ultra low power use. Huge leap toward brain scale AI.
Jan Vidic retweeted
Are you ready for one more computing breakthrough? - the 5th in the last 24 hours.
Breakthrough artificial neuron mimics the human brain, tiny memristor design could revolutionize AI efficiency.
Scientists have created a spiking artificial neuron powered by a diffusive memristor, a nanoscale component that behaves like a real biological neuron.
This revolutionary 1M1T1R (one transistor and one resistor) design packs all core neural functions - learning, firing, adapting, forgetting into a footprint the size of a single transistor.
Operating at picojoule energy levels, it鈥檚 MILLIONS of times more efficient than current AI chips. 馃憖
This breakthrough could enable neuromorphic chips with billions of brain like neurons, capable of real time learning, ultra low power use. Huge leap toward brain scale AI.
1h later...new massive computing breakthrough
New brain inspired chip learns on its own, no massive AI training needed
A team led by Dr. Joseph S. Friedman at The University of Texas at Dallas has developed a neuromorphic computing prototype, a brain inspired computer that learns and makes predictions using far fewer computations and much less energy than traditional AI systems.
Unlike conventional computers that separate memory and processing, neuromorphic hardware integrates both, mimicking how neurons and synapses in the brain adapt and learn.
The team used magnetic tunnel junctions (MTJs), nanoscale magnetic devices that strengthen connections between artificial neurons based on activity, following Hebb鈥檚 law ("neurons that fire together, wire together").
This breakthrough, published in Communications Engineering, is a big step toward energy efficient, self learning computers that can bring powerful AI capabilities to mobile devices without massive training costs.
Jan Vidic retweeted
This is the Singularity candle. In 2026, we鈥檒l be seeing it many times, each time following insane AI breakthroughs and a wave of new AI products (from tech giants) that will become increasingly frequent.
These AI innovations will completely change the world, and it will happen even before 2030.
BREAKING: Alphabet stock, $GOOGL, surges over +7% after reporting stronger than expected Q3 2025 earnings.
The AI Revolution is on fire.
Extropic Claims AI Could Run 10,000脳 More Efficiently With New Tech
Jan Vidic retweeted
10,000x more energy efficient AI soon?
Extropic claims that running Diffusion Transition Models (DTMs) on their Thermodynamic State Units (TSUs) could make generative AI up to 10,000x more energy efficient than today鈥檚 GPU based methods, according to their simulations.
Breakthrough Neuromorphic Chip Can Learn and Predict Without Big Data!
1h later...new massive computing breakthrough
New brain inspired chip learns on its own, no massive AI training needed
A team led by Dr. Joseph S. Friedman at The University of Texas at Dallas has developed a neuromorphic computing prototype, a brain inspired computer that learns and makes predictions using far fewer computations and much less energy than traditional AI systems.
Unlike conventional computers that separate memory and processing, neuromorphic hardware integrates both, mimicking how neurons and synapses in the brain adapt and learn.
The team used magnetic tunnel junctions (MTJs), nanoscale magnetic devices that strengthen connections between artificial neurons based on activity, following Hebb鈥檚 law ("neurons that fire together, wire together").
This breakthrough, published in Communications Engineering, is a big step toward energy efficient, self learning computers that can bring powerful AI capabilities to mobile devices without massive training costs.
馃毃 AI Leaps Forward: New Artificial Neurons Mimic Real Brain Cells馃馃殌
Another massive computing/AI breakthrough
Engineers create artificial neurons that think like real brain cells, big leap toward true AGI
Researchers at USC Viterbi School of Engineering have built artificial neurons that physically replicate how real brain cells process electrical and chemical signals, a historic step toward brain like computing and potentially AGI.
Powered by a breakthrough device called a diffusive memristor, these neurons use ions instead of electrons to compute, just like the human brain, enabling chips that are orders of magnitude smaller and more energy efficient than today鈥檚 silicon processors.
The new design, published in Nature Electronics, could revolutionize neuromorphic computing, making AI hardware that doesn鈥檛 just simulate thought but actually works like the human brain.
Jan Vidic retweeted
Another massive computing/AI breakthrough
Engineers create artificial neurons that think like real brain cells, big leap toward true AGI
Researchers at USC Viterbi School of Engineering have built artificial neurons that physically replicate how real brain cells process electrical and chemical signals, a historic step toward brain like computing and potentially AGI.
Powered by a breakthrough device called a diffusive memristor, these neurons use ions instead of electrons to compute, just like the human brain, enabling chips that are orders of magnitude smaller and more energy efficient than today鈥檚 silicon processors.
The new design, published in Nature Electronics, could revolutionize neuromorphic computing, making AI hardware that doesn鈥檛 just simulate thought but actually works like the human brain.
Jan Vidic retweeted
1h later...new massive computing breakthrough
New brain inspired chip learns on its own, no massive AI training needed
A team led by Dr. Joseph S. Friedman at The University of Texas at Dallas has developed a neuromorphic computing prototype, a brain inspired computer that learns and makes predictions using far fewer computations and much less energy than traditional AI systems.
Unlike conventional computers that separate memory and processing, neuromorphic hardware integrates both, mimicking how neurons and synapses in the brain adapt and learn.
The team used magnetic tunnel junctions (MTJs), nanoscale magnetic devices that strengthen connections between artificial neurons based on activity, following Hebb鈥檚 law ("neurons that fire together, wire together").
This breakthrough, published in Communications Engineering, is a big step toward energy efficient, self learning computers that can bring powerful AI capabilities to mobile devices without massive training costs.
Another massive computing/AI breakthrough
Engineers create artificial neurons that think like real brain cells, big leap toward true AGI
Researchers at USC Viterbi School of Engineering have built artificial neurons that physically replicate how real brain cells process electrical and chemical signals, a historic step toward brain like computing and potentially AGI.
Powered by a breakthrough device called a diffusive memristor, these neurons use ions instead of electrons to compute, just like the human brain, enabling chips that are orders of magnitude smaller and more energy efficient than today鈥檚 silicon processors.
The new design, published in Nature Electronics, could revolutionize neuromorphic computing, making AI hardware that doesn鈥檛 just simulate thought but actually works like the human brain.