
OpenAI debuted personal finance tools via Plaid for Pro users. AI startups generate ~$80B in annualized revenue, with Anthropic and OpenAI capturing 89%. ArXiv cracks down on AI slop, Apple's Siri relaunch may still be a beta, and SF vibes are frenetic.
Loading summary
A
Study and play come together on a Windows 11 PC and for a limited time, college students get the best of both worlds. Get the unreal college deal Everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass ultimate with a custom color Xbox wireless controller. Learn more@windows.com studentoffer while supplies last ends June 30th terms at aka mscollegepc.
B
Welcome to the Techbrew Ride Home for Monday, May 18, 2026. I'm Brad McCullough.
C
Today, OpenAI debuted personal finance tools via
B
Plaid for pro users. The big are getting richer in the
C
AI race and pulling away from the
B
rest of the industry.
C
Archive is cracking down on AI slop,
B
Apple's Siri relaunch may still be in beta, and the vibes around AI are getting feverish everywhere.
C
Here's what you missed today in the world of tech.
B
Today's episode is brought to you by Doppl. Social Engineering attacks Don't bother to knock.
C
They slip right into your inbox, your
B
phone, or on websites, instead pretending to be a harmless internal email or a normal text message until it's too late. Doppel sees right through the disguise.
C
Their AI native platform trains your team
B
to recognize threats like deepfakes, bad links, and impersonation attempts before they can actually cause any damage. Doppel strengthens team resilience by giving employees the tools and defenses they need to protect themselves from increasingly sophisticated social engineering threats. It's kind of like having a security team that has eyes everywhere, and their
C
digital risk protection takes it one step
B
further by keeping an eye on every channel to connect patterns and shut them down fast. Invest in social engineering defense. Learn more at D O p p
C
e l.com that's.o p p e l
B
dop Late last week, OpenAI debuted personal finance tools for US ChatGPT Pro users,
C
partnering with Plaid to give access to more than 12,000 financial institutions to analyze spending and more. Quoting TechCrunch users can connect over 12,000
B
financial institutions, including Schwab, Fidelity, Chase, Robinhood, American Express and Capital One. Once users connect these accounts, they will see a dashboard of their portfolio, performance, spending, subscriptions and upcoming payments.
C
The new product comes just one month after OpenAI acquired the team behind personal finance startup Hero, which was backed by firms like Ribbit, General Catalyst and restive.
B
In April, OpenAI said that the Hero team's expertise and finance was useful in launching this product, but didn't specify if
C
the entire feature was built by them.
B
OpenAI users can access the tool by
C
selecting Get Started in the Finances option
B
in the sidebar or typing at Finances,
C
Connect my Accounts and a chatgpt conversation
B
Once users do that, the chatbot will guide them about linking accounts through Plaid.
C
The company said it plans to support Intuit soon, which would enable analysis such as the impact of a stock sale on taxes or the odds of a credit card approval.
B
According to OpenAI, more than 200 million users already ask financial questions to ChatGPT every month.
C
The company also noted that the new GPT 5.5 model is stronger at reasoning with context, which is crucial for answering finance related questions. The company said it worked with Franklin
B
Finance experts to create a benchmark for
C
the model to improve on personal finance questions.
B
With the new financial tool integration, users can get detailed answers to questions such
C
as I feel like I've been spending more recently, has anything changed?
B
Or Help me build a plan to
C
be ready to buy a house in
B
my area in the next five years. Users can go to Settings, then Apps, then Finances to remove connections to certain accounts if they want.
C
Once they disconnect a service, the sync data will be removed from ChatGPT in 30 days.
B
What's more, users can also view and delete financial memory from the Finances page.
C
Generalized chatbots are designed to answer anything
B
leading people to ask questions about data
C
sensitive topics such as health, finance and personal life. AI companies are realizing this and making specialized products for these sectors. Both OpenAI and Anthropic have launched health related tools.
B
Earlier this month, Perplexity launched its own
C
financial research product based on its computer agent. OpenAI said its personal finance tools will be available on ChatGPT on the web and on iOS for pro users.
B
It noted that based on the feedback from these users, it wants to improve
C
the product before making it available to plus users. End quote.
B
Archive has had enough of the AI
C
slop quoting the Verge arXiv, a popular
B
platform for preprint academic research, is taking a new step to attempt to reduce the volume of papers that include AI slop. If a paper has quote incontrovertible evidence that the authors did not check the
C
results of LLM generation, such as hallucinated
B
references or meta comments Left by an LLM, authors will be banned from arXiv for a year. According to Thomas Dietrich, Archives Section Chair of its Computer Science section, future archive submissions will also have to be accepted at a reputable peer reviewed venue. Here's what he said on X Quote Attention at Archive authors. Our code of conduct states that by
C
signing your name as an author of
B
a paper, each author takes full responsibility for all its contents, irrespective of how the contents were generated. If generative AI tools generate inappropriate language, plagiarized content, biased content, errors, mistakes, incorrect references, or misleading content, and that output is included in scientific works, it is the responsibility of the authors.
C
We have recently clarified our penalties for this.
B
If a submission contains incontrovertible evidence that the authors did not check the results of LLM generation, this means we can't trust anything in the paper.
C
The Penalty is a one year ban
B
from arXiv, followed by the requirement that subsequent ARXIV submissions must first be accepted accepted at a reputable peer reviewed venue.
C
Examples of incontrovertible Evidence Hallucinated References Meta Comments from the LLM Here's a 200 word summary. Would you like me to make any changes?
B
The data in this table is illustrative.
C
Fill it in with the real numbers from your experiments, et cetera.
B
Authors can appeal ban decisions, Dietrich told 404 Media. He also noted that this policy will only apply to cases of incontrovertible evidence, and that our internal process requires first a moderator to document the problem and then for the section chair to confirm
C
before imposing the pen penalty.
B
Last year, Arxiv also updated its policies to reduce AI slop by only allowing computer science review articles and position papers to be published if they have been
C
peer reviewed and have been accepted at a conference or a journal. The advent of large language models have
B
made this type of content relatively easy to churn out on demand, and the majority of the review articles we receive are little more than annotated bibliographies with
C
no substantial discussion of open research issues, arxiv said at the time. End quote Foreign.
B
Suggests that there is a widening revenue gap between Anthropic OpenAI and the rest
C
of the entire AI industry. Quoting the Information A group of 34
B
leading startups in the AI sector, including Anthropic and OpenAI, is generating nearly $80 billion in annualized revenue, or 6.6 billion
C
per month, from selling AI applications or access to the models that power such apps, according to the Information's generative AI database. That's up 112% from just six months ago.
B
Investors at firms like Sequoia Capital have made the bet that the vast majority of software value in the current AI era will be generated by the top developers of advanced AI models rather than by developers of pure AI apps.
C
That's because almost all the AI application
B
companies in our analysis depend in part or almost entirely on models From Anthropic and OpenAI, the smaller startups reliance on Anthropic and OpenAI could be a problem
C
because those two leaders have been developing new kinds of models or versions of their products targeting specific industries or white collar work roles. See the earlier segment putting themselves in
B
direct competition with their startup customers.
C
That could eventually make it harder for those startups to keep up their growth.
B
That dynamic could help explain the frenetic funding environment for so called AI NEO labs that are seeking to leapfrog anthropic and OpenAI by developing new kinds of AI models. The $80 billion annualized sales figure is remarkable as few of the companies existed or generated revenue before 2023. Anthropic can claim about half amount as
C
its Revenue recently surpassed OpenAI's on the strength of its AI for coding tasks and other white collar work.
B
The 32 AI model or application startups not named Anthropic or OpenAI may not
C
be growing as fast as the leaders,
B
but they aren't far behind in that
C
regard and continue to hit new revenue milestones largely through sales of subscriptions and some usage based Feees.
B
Since December, three of those companies crossed
C
$500 million in annualized sales, joining coding
B
app Cursor in that club, according to our analysis. The three newbies are search firm Perplexity, voice AI provider ElevenLabs and coding app Cognition.
C
Our analysis doesn't include the infrastructure startups that rent out Nvidia AI servers.
B
Now there is some double counting going on.
C
The four aforementioned startups that are generating more than $500 million in annualized sales, as well as the other companies on
B
the list likely collectively pay OpenAI and anthropic billions of dollars a year for
C
models to power their products. Almost all of the company's products are powered by Nvidia servers, helping the chip designer generate the vast majority of profits so far in the AI boom, Nvidia has partly returned the favor, investing capital
B
in at least 13 of the 34 companies on our list, including about $40 billion in OpenAI and anthropic combined.
C
The capital helps the companies keep renting Nvidia servers.
B
Our analysis doesn't cover public companies that are also generating revenue from AI applications and selling models they develop themselves. Microsoft, for instance, has generated at least $6 billion in actual revenue from selling Copilot branded apps, power in part by
C
OpenAI models that it uses for free.
B
Google sells a lot of Gemini AI
C
models and Gemini powered features in enterprise apps, and AI Answers appear to be boosting its core search engine and by extension, the search advertising business.
B
A dozen major enterprise software firms such
C
as Salesforce and ServiceNow are collectively on pace to generate billions of dollars a
B
year selling AI powered features and agents in their products. There are also major customers of Anthropic and OpenAI, both for internal use and
C
to power their products. They sell these AI buyers have been enduring recent price hikes from Anthropic End quote.
B
There's a lot to navigate in the economy right now, and small business owners like you are really feeling the effects. You can't control interest rates or tariffs, but there is one thing you can control, and that's how efficient your business is. Gusto can help Automating payroll and HR with Gusto is one of the fastest
C
ways to cut friction and focus on
B
what actually moves the needle. Gusto is online payroll and benefits software built for small businesses. It's all in one remote, friendly and incredibly easy to use so you can pay, hire onboard and support your team from anywhere.
C
Unlimited payroll runs for one monthly price.
B
Try Gusto today@gusto.com Brewster and get three months free when you run your first payroll. That's three months of free payroll@gusto.com Brew Father's Day is right around the corner. Celebrate the dads in your life with Viori.
C
Their Father's Day gift guide features the
B
newest and best selling styles dad will want to wear on repeat. From the fairway to the airport to the gym, Vuorie's everyday staples are ready for wherever the day takes him. With fitness Go to's Performance Essentials and Travel Ready layers, you can gear him up for his next adventure in one place. Get 20% off your first order at viori.com techbrew that's vuori.com Techbrew exclusions apply. Visit the website for full terms and conditions. I keep getting this weird Spidey sense that Apple's Siri AI rollout is going
C
to underwhelm once again. Quoting Mark Gurman, what's old is new again.
B
Apple historians may remember that when the company rolled out the original Siri back in 2011, it did so by branding the voice assistant as a beta, a
C
label meant to denote that it's not a fully baked product and may include some issues. The company ended up dropping that nomenclature in 2013 as part of iOS7. But let's be honest, Siri has still
B
felt like a beta or even rougher
C
for the better part of 15 years.
B
Anyways, back to the point. Test versions of iOS 27 within Apple use this label for the new Siri and include a toggle to leave the the Siri beta.
C
Given that we're only a month out
B
from wwdc, there is a strong chance
C
that this approach will be used in developer beta versions and even when iOS27 ships this fall.
B
This all means that even after a two year delay, the revamped Siri was
C
supposed to arrive in 2024, the company
B
could still brand the new features as unfinished.
C
It's a far cry from what we're seeing in the Android universe. End quote. Which yeah, Google I O's keynote is tomorrow and expect a metric ton of AI announces there.
B
Let's end with a couple of vibey
C
stories around the Vibes of AI at the moment. In a Times Opinion piece, a college
B
senior at Stanford describes how AI has changed classwork and college generally.
C
Cheating using AI has, according to them,
B
become omnipresent, with students fudging just about everything. Students were probably the earliest wide scale adopters of AI.
C
After all, it was far and away the quickest route to an A. When I took CS107, the only viable
B
way for people to cheat was to
C
seek out a student who'd gone through
B
the class before and beg for solutions
C
to the notoriously difficult problem sets.
B
There was no alternative to putting in a large amount of work, even if one did obtain the answers from another student.
C
Engaging, by the way, in a social act if nothing else.
B
The students I knew who did this
C
still spent hours sculpting their stolen code so as not to be caught.
B
Few cheated in this most overt fashion back then, but a month later, any student could instead turn to a chatbot,
C
plugging in a prompt, alone in a dorm room, and mindlessly regurgitating the result.
B
I remember the first time I used
C
it, feeling an immediate sense of guilt. A friend recently told me. Now it's just normal.
B
Half of the laptops in any lecture
C
seem to be open to ChatGPT or Claude. In the beginning, experimenting with models was a pastime for the nerds showing off the the early access you got to the next frontier. Large language model was a status symbol,
B
and people would come pleading for your
C
authorization keys to try it out for themselves.
B
In just a few short years, however, AI has become a fact of life.
C
It's all we talk about, my ancient Greek art history professor remarked recently. End quote.
B
And
C
on X had a tweet that
B
got people either schadenfreuding or popcorn eating
C
or mocking or all three over the weekend. I'm just going to read it.
B
Quote the Vibes in San Francisco feel pretty frenetic right now. The divide in outcomes is the worst I've ever seen. Over the last five years a group
C
of around 10,000 people employees at Anthropic OpenAI XAI Nvidia Meta TBD founders have hit retirement wealth of well above $20 million back of the envelope AI estimation.
B
Everyone outside that group feels like they
C
can work their well paying jobs but sub 500,000 dol a year for their
B
whole life and never get there. Worse yet, layoffs are in full swing.
C
Many software engineers feel like their life skill is no longer useful. The day to day role of most jobs has changed overnight with AI as a result.
B
Number one, the corporate ladder looks like the wrong building to climb.
C
Everyone's trying to align with a new set of career paths. Should I be a founder? Is it too late to join Anthropic or OpenAI? Should I get into AI?
B
What company stock will 10x next?
C
People are demanding higher salaries and switching jobs more and more.
B
Number two, there's a deep malaise about
C
work and its future. Why even work at all?
B
For peanuts?
C
Will my job even exist in a few years? Many feel helpless. You hear the permanent underclass conversation a lot, especially from young people. It's hard to focus on doing good work when you think man, if I joined Anthropic two years ago I could retire.
B
Number three, the mid to late middle managers feel paralyzed.
C
Many have families and don't feel like they have the energy or network to just start a company. They don't particularly have any AI skills. They see the writing on the wall. Middle management is being hollowed out in many.
B
Number four, the rich aren't particularly happy either. No one is shedding tears for them and rightfully so. But those who have quote made it
C
experience a profound lack of purpose as well.
B
Some have gone from less than $150,000
C
to greater than $50 million in a few years with no ramp.
B
It flips your life plans upside down.
C
For some, comparison is the thief of joy. For some they escape to New York City to just live life. For others they start companies just cause often to win status points.
B
They never imagined that by age 30
C
they I once asked a post economic founder friend why they didn't just sell the company and they said and do what?
B
Right now everybody wants to talk to
C
me and if I sell I will only have the money.
B
I understand that many reading this scoff
C
at the champagne problems of the valley. Society is warped in this tech bubble. What is often well off anywhere else in the world is being average here
B
Unlike many other places, tenure, intelligence and hard work can be loosely correlated with
C
outcomes in the Bay. Living through a societally transformative gold rush in that environment can be paralyzing.
B
Am I in the right place?
C
Should I move?
B
Is there still time left?
C
Am I going to make it? It psychologically torments many who have moved here in search of success. Ironically, a frequent side effect of this torment is to spin up the very products, making everyone rich in hopes that you too can vicode your path to economic enlightenment. End quot.
B
So tomorrow is Google I o.
C
But as I mentioned earlier in the
B
show, there will be so much announced
C
and it will get started only at
B
1pm Eastern that I think the only sensible thing to do is delay talking
C
about IO until Wednesday's show.
B
That way things can settle out and
C
we'll have a better sense of what was important and what wasn't. As opposed to me just rattling off everything announced in chronological order without having the visibility to emphasize the importance of
B
stuff over the unimportant.
C
And also, as I'm putting the final touches on this episode, the jury just came in on the Musk v. OpenAI trial, so we need to make room for that. Talk to you tomorrow.
D
Did you know if your windows are bare, indoor temperatures can go up 20 degree. Get ahead of summer with custom window treatments like solar roller shades from blinds.com and save up to 45% off during the Memorial Day Early Access sale. Whether you want to DIY it or have a pro hand handle everything, we've got you free samples, real design experts and zero pressure. Just help when you need it. Shop up to 45% off site wide right now during the Early Access Memorial day sale at blinds.com rules and restrictions apply.
Episode Theme:
This episode dives into OpenAI’s new push into personal finance for ChatGPT Pro users, AI’s growing dominance (and the increasing gap between mega-players and the rest), arXiv’s crackdown on AI-generated "slop", an underwhelming preview of Apple’s forthcoming Siri overhaul, and the often-unspoken vibes of anxiety, malaise, and transformation rippling through Silicon Valley and the wider workforce due to accelerating AI trends.
Key Points:
Sample User Scenarios:
Quote Highlight:
"OpenAI users can access the tool by selecting Get Started in the Finances option in the sidebar or typing, '@Finances, connect my accounts,' in a ChatGPT conversation. Once users do that, the chatbot will guide them about linking accounts through Plaid." — (02:45)
Notable Details:
Key Points:
Notable Quote:
"Our code of conduct states that by signing your name as an author of a paper, each author takes full responsibility for all its contents, irrespective of how the contents were generated." — Thomas Dietrich, arXiv Section Chair (05:22)
Key Points:
Quote:
"Almost all the AI application companies in our analysis depend in part or almost entirely on models from Anthropic and OpenAI..." (07:44)
Key Points:
Notable Quote:
“Apple historians may remember… the original Siri [was] branded as a beta… But let’s be honest, Siri has still felt like a beta or even rougher for the better part of 15 years.” — quoting Mark Gurman (12:16)
Informative, concise, and slightly sardonic, with a pulse on both the product and emotional shifts in the tech world. The episode combines fast-paced industry news with deeper reflection on tech culture’s mood swings during a period of rapid AI disruption.