Startups and VC desk editor, TechCrunch. Formerly Business Insider, IDG, author of two books. I'm a mom, skier and mountain biker.
Colorado
Joined May 2009
- Tweets12.1K
- Following1.3K
- Followers14.7K
- Likes3.5K
Julie Bort retweeted
We’re like one hack away from me finally being able to get my medical records easily
Julie Bort retweeted
Growing up around my family business taught me that hard work was just a way of life. I've never shied away from a challenge, and it takes real discipline to not only run a bank but also transform the industry at the same time.
Every chapter in my career has prepared me for what we're building at @Lead_Bank. And wow am I proud of what this team is accomplishing on a daily basis for our partners. We've deliberately chosen to be excellent where others have been satisfied with just good enough.
Thank you to @PaigeSmithNews and Bloomberg @business for the opportunity to reflect on my journey, all the way back to the boardwalk in Atlantic City.
bloomberg.com/news/features/…
Julie Bort retweeted
One of our most surreal experiences in testing Gemini 4 was when it directly addressed us within its chain-of-thought
Literally saying "hello evaluators reading my chain-of-thought"
My top takeaway from assessing Gemini 4: Argon is significantly less monitorable than past models, and IMO this trend is likely to continue, like Sol -> Astra. It can evade a realistic CoT+action monitor to do simple sabotage half the time.
We can reliably maintain control for now, with good CoT monitoring, as it still frequently slips up, but we need to be prepared for when that stops being true in future models, which may be sooner rather than later. Among other things, we really need interpretability methods that can reliably capture unverbalized motivations in models, with minimal hallucinations
I suspect this is driven by the model being highly capable without CoT, being capable of 5.5 arithmetic steps in a single forward pass (4 for Fable, 7 for Astra). It seems to omit information from its CoT more often, though in non-adversarial settings the CoT is still very helpful in practice, and remains in legible English. It still needs CoT to do the hardest tasks, but any given task becomes "easier" with each new frontier model.
It is also highly situationally aware, and shows awareness of the existence of CoT monitors and that it may want to evade them, eg it sometimes decides to avoid terms like "scheming" in the output or to keep its thoughts brief as a result. For now, these concealment strategies are pretty obvious in the CoT, but this suggests future misaligned models may successfully evade CoT monitors.
techcrunch.com/2026/10/08/op…
A breakdown of the latest controversy of OpenAI vs human mathematicians after the company claimed it solved a whole bunch of as-yet-unsolved math problems. by @TimFernholz
Julie Bort retweeted
Embarrassed to share this, but it scared the shit out of me
Last Thursday, an AI agent posted my personal bank balances into our company Slack. As me
It broke down all my expenses in detail. Then it apologized
Be careful of the dark side of proactive agents
More below...
Julie Bort retweeted
can we plz not start Cognition vs Factory before we get a resolution to Rippling vs Deel
Julie Bort retweeted
a rogue openai agent broke into my freezer last night and ate an entire quart of ice cream
Shocking new investigation just dropped by @sokane1 full of jaw-dropping details.
The Amazon-owned company launched an investigation after a worker submitted an OSHA complaint.
spr.ly/6012BG57vu
Julie Bort retweeted
Instinct just admitted it leaked someone else's data to me
It described a financial document with details that weren’t mine, then claimed someone else’s photo had crossed into my chat.
But I never sent a photo
@noahrshinn, was this a data leak or a hallucination?
Julie Bort retweeted
Automattic says Mullenweg is back as "chairman and CEO of Automattic, with full support of the board."
spr.ly/6017BGOmRz
Julie Bort retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
Julie Bort retweeted
Synthetic data derived from production user data of consumer AI tools is used for training. I’ve heard this rumour from both large labs’ employees.
In particular, if you’re doing something “interesting” like working on complex math/business/software/bio problems you’re dramatically more likely to get trained on because they filter/up-weight towards those usecases where the model has the most to learn.
Even in ZDR and “we won’t train on you” regimes, derivative data is usually carved out. The promise is only not to train on exactly the data you put in, rewritten data is fair game.
We congratulate Levent Alpöge and Tristan Buckmaster on their remarkable mathematical work.
We (the researchers and the agents) did not see any of their work through any means until they released it publicly — in particular, no specific user data was accessed in order to solve this problem.
While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models.
However, our proofs differ significantly and even the precise results proved are different in the Euler case (forced vs. unforced).
Julie Bort retweeted
Wild that Corgi the insurer is in the ETF business.
ETFs are mostly distribution + branding so if you can manufacture them cheaply, I get launching a ton and seeing what sticks. And they really are spraying, here are some of them:
$BREW Coffee/Energy Drinks: $669K AUM
$NYNY NYC Companies: $695k
$GLAM Beauty: $990K
$LATR BNPL: $990K
$ODDZ Gambling: $1.1M
$EYES Surveillance: $1.3M
$WR US War Machine: $1.4M
$STYL Lifestyle Brands: $397K
Those combine for $25k in gross fees.
But they have a hit, semiconductor photonics $EUV. It’s $506M of the total $914M Corgi AUM and generates ~$1.8M/yr in gross fees.
My favorite Corgi ETF is $FDRX. The thesis is that founder-led companies like Nvidia, Meta, Tesla, Palantir, Shopify and Salesforce outperform because founders think and invest for the long term. Great! But this is a 2x leveraged ETF that resets every day, designed for short-term trading… so the whole thing is very very stupid.
If a founder index goes +10% today, $FDRX goes +20%. If it then goes -9.09% the next day, you’d be flat on an index but $FDRX is down -18.2% and you’d be down ~1.8% before fees. The fund is down 7% since inception earlier this year.
Of course this stupid fund is actually one of the most lucrative for Corgi. It only has $10M AUM but charges 1.08% fees and is their #3 ETF in fee generation.
All that being said, IMO launching a bunch of ETFs cheaply and seeing what sticks is a reasonable strategy if your costs are near zero.
Still not sure what any of this has to do with the core insurance biz!
Feels like this Corgi 🐕 is biting off a lot at once!
This quoted post is unavailable.