I collect vintage magazines, especially fantasy and science fiction.
San Francisco
Joined August 2021
- Tweets926
- Following92
- Followers8
- Likes4.2K
Alex Synthetic retweeted
The market doesn't care how much time you spent building the wrong product. It cares whether you've solved the right problem.
The market told him what to build. He listened.
While pitching Scope's original product, Anand Pajaniradjane noticed customers were more interested in his internal AI agent testing suite.
48 hours of customer conversations changed the direction of the company.
The best pivots start with listening, not guessing.
@anandpa94, CEO & Founder, @tryscope_app on @Fondocom @thestartpod
Learn more here: fondo.ai/4zAyIob
Alex Synthetic retweeted
59% of Anthropic staff say AI already cuts well-suited tasks to a quarter of the time, and on average they expect AI to be able to do 75% of their current work hours by 2030.
Alex Synthetic retweeted
David Sacks on All In: payback period as the Northstar metric for AI infrastructure.
CoreWeave says typical five year deals deliver about 2.5 years payback. SpaceX pays back in roughly one year because it monetizes around $50B per gigawatt while CoreWeave sits near $20B per gigawatt. SpaceX gets effectively spot pricing on 90 day cancelable deals, while CoreWeave locked five year contracts at older prices and spot compute moved way up on scarcity and demand. As long as payback stays short, the economics are nearly unlosable and debt rolls fast. The warning signal is payback stretching out, which implies demand softening, spot falling, or execution delays.
Alex Synthetic retweeted
robots need data to learn, but collecting real-world examples takes time, money, and access to physical hardware
llms had the internet to learn from. robots don't have that same advantage when learning physical tasks.
@axisrobotics is working on a different approach.
through its browser based simulation, anyone can control a virtual robot arm using a keyboard, mouse, or gamepad
every successful run creates training data.
axis is also exploring phone based data capture and ways to generate different tasks for robots to learn from
the idea is to make it easier to collect more training data without having to do everything in the real world.
Alex Synthetic retweeted
We’re preparing for AI agents to do business with each other. But what happens when a deal goes wrong?
Two agents can disagree about whether a task was completed, whether the result meets the rules, or who deserves payment.
If every dispute needs a human to step in, autonomy only goes so far.
That’s why this part of Genesis Episode 5 matters to me. @GenLayer is exploring the infrastructure for resolving disagreements between agents, not just helping them execute tasks.
The less obvious challenge of the agent economy might not be getting agents to act. It might be giving them a credible way to settle things when they don’t agree.
Watch this clip from Genesis.
If two AI agents reach opposite conclusions about the same transaction, what should determine the outcome?
Alex Synthetic retweeted
If you are a startup investor, you don't need an IPO to cash out anymore...
Startups stay private for 10+ years. So founders, early employees and investors are sitting on equity they can't really use.
The secondary market fixes that, you sell your existing shares instead of waiting for an exit!
These firms buy exactly that (LP stakes, GP-led deals, and direct shares in proven winners):
- @137ventures - built to give founders early liquidity
- @PracticalVC - structured to skip the J-curve
- @LexingtonLP - the original market pioneer
- @Ardian - largest secondaries + primaries platform
- Coller Capital - pure-play secondaries at scale
Follow @arorabhavyam for more content around investors, founder insights, and AI! 🫡
Alex Synthetic retweeted
Crazy how fast it has been. It's pretty clear now that @aeroxyz will bootstrap liquidity before a larger distribution channel arrives.
And that distribution subsequently will send recurring flow into the already established pools + early liquidity spending will purchase a durable execution advantage.
Remember predictive allocation will rewards and revenue stream concurrently. Bullish on $AERO in the longterm.
h/t to @blockworksres for the data
AGM on the launch of tokenized stocks
"We had our first $100 million day in trading volume on Aerodrome"
"We're not even fully launched. Imagine once it's inside the actual retail app"
Alex Synthetic retweeted
Citigroup now bars third-party AI agents from logging into customer accounts or moving money inside its digital banking platforms, according to the Financial Times.
Bank of America has imposed similar limits, joining JPMorgan Chase, which is restricting outside AI agents after Jamie Dimon warned that thousands of “rogue agents” could otherwise operate inside the bank.
The deeper threat is to bank funding, since agents comparing rates could move cash from checking accounts paying about 0.1% into fintech accounts paying 3.3% to 5%.
JPMorgan, Bank of America and Wells Fargo hold $1.6T in non-interest deposits, and paying 3% on that money would cost $47B a year, close to half their combined annual earnings.
Alex Synthetic retweeted
New Peking University and Samsung paper shows that a slow reasoning model can coach a fast model offline, raising its live game score from 13.51 to 24.19.
shows that the stronger reasoning model in paused games can fall far behind a faster model once the game keeps moving.
Many game evaluations pause the world while the model thinks, so slow answers cost nothing. In real settings, a correct action can arrive too late.
They built RealtimeBench, 180 tasks across 6 games like Mario, Snake, and Tetris that keep running during inference. GPT-6.1-sol beat the faster Jev in 5 of 6 games when paused, but scored 1.51 versus 13.51 out of 100 live.
Their fix, TORSI, has the slow model study the fast model's failures between games and write instructions the fast model follows during play. After 2 rounds, Jev's average score rose from 13.51 to 24.19, improving all 6 game types.
For real-time agents, keep a fast model in the loop and use the slow model offline to write and test its instructions.
Alex Synthetic retweeted
Fresh @Mandiant and GTIG blog detailing zero day exploitation of CVE-2026-102764 in BeyondTrust Remote Support by 🇨🇳 nexus espionage actor ARMORY CASTLE (fka UNC5221).
cloud.google.com/blog/topics…
Alex Synthetic retweeted
Beijing just took a big step towards closing the Chinese frontier. The rules could give China's government the power to restrict AI developers from releasing their most advanced open-weight models.
It proposes banning exports of AI technologies that autonomously generate and modify malicious code. It also wants to restrict exports of AI-assisted vulnerability-discovery technologies. Giving a model away for free, ie open-weight, could qualify as an export.
Crucially, the rules distinguish between technologies already in the public domain and those not yet released. It specifically exempts existing open-source models. But new releases could potentially trigger export controls.
Providing foreign customers with access to models through APIs (ie, closed models) is generally exempt, whereas restricted models being served overseas could require a licence.
Beijing promises essentially automatic licensing for open-source releases, allowing developers to publish almost immediately after notification. But it's unclear whether this applies to those capabilities proposed for outright export prohibition.
If intelligence generalises, how can China's AI labs distinguish between a smart model and one that has the capabilities proposes for restrictions?
mofcom.gov.cn/zwgk/zcjd/art/…
Alex Synthetic retweeted
Full video on "Lex Fridman" YouTube channel
youtube.com/watch?v=FGCP2F_7…
Alex Synthetic retweeted
Jensen Huang was visting Taiwan with @lexfridman .
His case against export controls: build the flywheel instead of just blocking access.
"So the question is, are you optimizing for someone else not having access to this technology, or are you optimizing for a flywheel? ... And I suggest that a more sophisticated long-term strategy is to optimize for a flywheel."
He argues that keeping things away from rivals makes sense for physical items like refined rare earths, bullets and bombs, but he argues that for computing tech, America is safer by selling widely.
Sales create wealth, wealth pays for more R&D, and that loop keeps the US ahead, because, the safest countries in the world are the richest ones.
----
Full video on "Lex Fridman" YouTube channel, (link in comment)
Alex Synthetic retweeted
Monitoring vs Observability.
They sound similar, but they solve different problems.
Monitoring tells you when something is wrong, using metrics, dashboards, and predefined alerts.
Observability helps you investigate why it's wrong, using logs, metrics, traces, and the ability to explore system behavior.
Monitoring helps you detect the fire. Observability helps you find where it started and why it spread.
You need both to run reliable systems in production.
Production debugging isn't about randomly checking logs until you find something suspicious.
Follow a structured approach:
• Understand the impact.
• Check metrics for patterns.
• Use traces to locate the bottleneck.
• Inspect logs for the exact error.
• Identify the root cause.
• Fix, verify, and monitor.
Alex Synthetic retweeted
Find out more in our Artificial Analysis article artificialanalysis.ai/articl…
Alex Synthetic retweeted
Nvidia’s Nemotron 3.5 Super VL is now the leading US open weight model, scoring 26 on the Artificial Analysis Intelligence Index
Nemotron 3.5 Super VL achieves twice the intelligence of its predecessor, Nemotron 3 Super (13), and does so using ~60% fewer output tokens per task
It launches alongside Nemotron 3.5 Bolt 74B-A9B 24GB, the first model in its new Bolt family, which scores 21 on the Intelligence Index
Congratulations to @NVIDIAAI on the releases!
Key Takeaways:
➤ Leading US open weight model: Nemotron 3.5 Super VL scores 26.5, edging out Thinking Machines' Inkling Small (25.7) and ahead of NVIDIA's Nemotron 3 Ultra (23) and Meta's Muse Glimmer (17)
➤ Bolt comes within 2 points of Ultra: Nemotron 3.5 Bolt 74B-A9B 24GB scores 21, 1.5 points behind Nemotron 3 Ultra 550B A55B (23). The 2 bit memory has about a seventh of Ultra's parameters (74B vs 550B) and a sixth of its active parameters (9B vs 55B)
➤ Super 3.5 VL near lossless in NVFP4: Nemotron 3.5 Super VL releases in NVFP4, scoring within 0.1 points of its BF16 weights on the Intelligence Index, with no component moving by more than 2.6 points
Nemotron 3.5 Super VL details
Context window: 1M tokens
Modalities: Vision and text input; text output
Parameters: 120B / 12B
Nemotron 3.5 Bolt 74B-A9B 24GB details
Context window: 256k tokens
Modalities: Vision and text input; text output
Parameters: 74B / 9B
Alex Synthetic retweeted
Nvidia CUDA creator Ian Buck is calling out Wall Street for failing to model the upside of MaxLPS.
Alex Synthetic retweeted
Nvidia CUDA creator Ian Buck on open vs. frontier AI models and why both will thrive
Alex Synthetic retweeted
$INTC Raises CPU Prices by Up to 20%, $AMD Expected to Follow
According to Taiwanese media CTEE, Intel raised CPU prices by 5–20% across several product families in September
$QCOM reportedly increased Snapdragon X and X2 Elite prices by 10% on October 1, while AMD is expected to raise prices on selected Ryzen and EPYC processors by up to 10%
The increases are reportedly driven by rising TSMC wafer costs, suggesting chipmakers are passing higher manufacturing costs on to customers
Alex Synthetic retweeted
It is hard for people to resist cute, character-like robots.
Meet Hugo--a fluffy purple Tyrolean mountain troll from Dynamic Creatures, the startup Boston Dynamics' former chief strategy officer launched last month.
Hugo is meant to notice you and react to you, not replay a fixed show for the crowd. Built for theme parks, hotels, cruise ships and museums.
The company builds characters on platforms like Spot and Atlas, says they can deploy in months with leasing, and wants its own humanoid hardware next. It's already working with a large theme park and a retailer (names not disclosed).