FastForward #74: AI makes critical thinking more important than ever
Hi everyone. The summer is rolling along and I'm happy to be back home. I want to thank each of you for reading my newsletter every week. Please share this week’s edition with a friend, and encourage them to subscribe. It really helps.💌 Sign up here.
ForwardThinking 🤔
AI makes critical thinking more important than ever
I was at a party last weekend and had a couple of conversations with college professors about their use of AI. I was curious how they were dealing with the tools, whether they were embracing them or blocking them. These two were at opposite ends of the spectrum, though interestingly for the same reason. They wanted to encourage their students’ critical-thinking skills, either with the help of the models or without them.
Students develop the basic intellectual building blocks in elementary and high school. College, in my view, is where you learn how to learn and develop independent thought and judgment, regardless of your major. AI seems to offer the siren's song of a shortcut, but in reality, it makes those advanced skills more important than ever.
The anti-AI professor, who works at a Midwestern college, said something that stuck with me. "We are skipping a critical step." What she meant was that students have to learn to assess the value of the answers that the models give them, rather than accepting them at face value. The question is how to do that when it seems so easy just to ask and report.
Her response was to ban it completely, not allowing any electronics in her classroom.
The second professor, from a school on the East Coast, was a bit more pragmatic. The professor encouraged students who wanted to use AI, while letting those who didn't for environmental or other reasons, do the work the old-fashioned way.
Last year, he gave them a bunch of reading assignments, then told them to ask ChatGPT some questions about the content and critique the answers. That worked until students started using Claude to critique ChatGPT. What's a professor to do?
They are both moving back to more oral work, more in-class essays, more blue book exams and group projects. It still leaves the question of how to teach research and writing, a critical academic exercise, while balancing AI (or not).
Getting those reps
Once those students graduate from college and get a job, they are going to need to get the same kind of critical training that only entry-level jobs can provide. This is where they begin to apply those fundamental skills they learned in college, and why a college education remains essential, especially in the age of AI.
As I was thinking about this, I came across this video from marketing guru extraordinaire, Robert Rose. He put into words what the Midwestern professor called, "skipping a step." He calls it formative debt, which he defined as "what accumulates when you skip the experiences that are supposed to shape who you are becoming." What does he mean by that?
Well, each of us, especially those who have been around the block a few times and accumulated some professional scar tissue, has developed a kind of judgment almost without realizing it. For me, it's what constitutes a good story. For Robert, what makes a good marketing campaign. The profession doesn't matter. You have to get your hard knocks before you get good at something. There is no shortcut.
As he said, "You see, doing the work, the friction-filled work, really doing it with all its friction. Well, it leaves a deposit in us, in our perception, in our judgment, in who we are becoming as a professional."
It's why we need an education. It's why we need entry-level jobs where we are allowed to make mistakes, to get knocked down, to feel the sting of failure. It's only through that bitter experience that we get good at something.
AI doesn't offer a replacement for that. On the contrary, it makes those experiences even more important because you have to be able to judge the output, push back when it's wrong, and ultimately find your own way in the world.
~Ron
What's new on the blog 📰
Why AI agents need a zero trust foundation
In part 2 of my three-part video interview with Zscaler's Swamy Kocherlakota, we talked agents, what they mean, how to get visibility into what they're doing and how to protect them as they operate inside an organization.
Kocherlakota talked about applying Zscaler's zero trust approach to agents:
"The number one question is: Who are you? What are you trying to do? What is the risk? What is your intent? And then policy allows you to say you're allowed to do that or not."
Note that the interview series is part of a paid engagement with Zscaler. My editorial standards and guidelines apply regardless. See my full disclosure statement at the bottom of the piece.
After IBM stock takes a beating, what comes next?
In my latest collab with Alex Wilhelm from the Cautious Optimism blog, my former TechCrunch colleague and I dig into IBM's disastrous second quarter earnings report, one so bad that CEO Arvind Krishna sent out a letter to shareholders before the report came out to soften the blow and let them know what was coming.
After that letter was delivered, investors reacted badly with a record one-day stock price drop of 25%. The question is, was this an over-reaction to one bad quarter or a more dangerous signal.
The startup unit inside Zscaler working to secure AI
Why would a large company build a startup unit within the broader organization? When I was at Zscaler Zenith Live 2026 in Las Vegas last month, I sat down to do a video interview with Swamy Kocherlakota, who runs such a unit inside Zscaler, to learn more about his role.
He said it was about generating speed and agility inside a larger organization that tends to move more slowly:
"AI is moving very fast, and customers' needs are moving fast. Every day there is a press release or an update. The question really is, how do you build products that allow our customers to protect AI and take advantage of AI."
Note that this interview series is part of a paid engagement with Zscaler. My editorial standards and guidelines apply regardless. See my full disclosure statement at the bottom of the piece.
Pega founder Alan Trefler thinks big AI is selling fear
Pega CEO Alan Trefler founded his company in the 1980s, and he's been around long enough to call out big AI when he sees fear mongering and manipulation.
"They're both trying to scare them into thinking it's all powerful, and also scare them into thinking that if they don't use it, they're going to fall behind."
I spoke to Trefler about how his company's workflow software has been designed to use AI strategically, rather than as the center of the system, an approach he believes will help them control costs and give customers more predictable bills.
Investor Jai Das on AI hype and the changing face of enterprise investing
I caught up with Sapphire Ventures co-founder Jai Das in April at HumanX. We talked about AI hype and the changing investment environment.
As the AI labs try to muscle into areas traditionally covered by Enterprise SaaS companies, he says there are still areas where enterprise investors can succeed.
"I'm in a little bit a better position than my colleagues who look more at apps because I do more on the infrastructure and software. So those layers, I think, are not going to get disintermediated by the models as much.”
News of the Week 📣
Figma has a killer quarter, Wall Street yawns
It seems as though Wall Street investors might have it out for SaaS stocks. No matter how well a company seems to do, the street finds a way to put a damper on it. But no Debbie Downer could take the shine off what Figma reported this week.
First the good news: Revenue was up a whopping 48%, which is pretty amazing for a company as mature as Figma. They beat estimates, raised their revenue guidance for the year, while keeping their profit projections intact – sounds pretty good, right?
I would say so, but what do I know? Instead of rewarding a seemingly great quarter, investors smacked Figma, which was down 15% overnight. How does this make a lick of sense?
Well, part of it is the high cost of implementing AI. As we've discussed previously, AI hardware is expensive, so even though Figma is delivering AI products, and customers are buying them, the cost of delivery is so high, it's eating into their operating costs.
While that revenue was up, it came at a cost with research and development costs up 101.5%, while doubling total operating cost over the prior year, per Reuters. That spending level has investors concerned (just as it does with the hyperscalers).
That will require more products and more revenue. My buddy Alex Wilhelm, writing on his Cautious Optimism blog (which is really good and you should subscribe) said it was because Figma had some products in beta, and were reluctant to project that revenue forward, even though it's probably coming.
Patience has never been a market virtue and Figma has to learn to balance competing investor demands around growth and costs. It's never easy.
Crunchbase: EU AI venture money has record first half
A new report released this week on European venture capital found that the region had its best half ever in terms of AI investments, by far, topping out at $23 billion in funding for the half, equalling the total raised across the entire year in 2025.
The report was released jointly by Crunchbase and Human X, the AI conference, which will be taking place in Amsterdam in September (and where I'll be moderating some panels). AI also represented a new high of 55% of overall funding in the region.
While the numbers pale in comparison to U.S. venture money, these figures show that AI is driving new venture growth beyond the U.S. borders. Consider that the U.S. pushed out $334 billion in the same time period, per Crunchbase. Of that, $217 billion went to two companies: Anthropic and OpenAI. Big shock.
Even when you subtract the two big AI labs' funding, it's still a huge gap of $117 billion versus $23 billion, but the EU still has to feel good about itself. It has a far less mature startup ecosystem than the U.S. and experienced 130% growth in AI startup investment in the half compared to H1 2025.
Bending Spoons buys Airtable for a fraction of its high-water private valuation
Airtable, once a sizzling startup with a valuation of $11.7 billion, was acquired this week by Bending Spoons, the Italian private equity firm hoovering up once-valuable SaaS companies. The price? An equity value of a mere $2.25 billion, and it's not even that good.
When you subtract the company's substantial cash hoard, the price drops all the way to $1.285 billion, per Cautious Optimism. That's a steep fall from grace, and it shows just how inflated those valuations were in late 2021, the peak of the ZIRP era.
The folks who got in that final round showed little return on their investment. Interestingly, there were institutional investors like Franklin Templeton, T. Rowe Price and J.P. Morgan, all investors who turn up when they sense an IPO is coming. It never came.
They invested in Airtable, a low-code application platform that looked like a souped-up spreadsheet but worked more like a lightweight database (one I use to track my FastForward profile interviews). In 2021, as I wrote on TechCrunch, low-code, automation and RPA were coming together and looked to be unstoppable…until generative AI and a SaaS valuation reset came along and basically upended the whole apple cart.
This week we saw the result. It was good software. It tried to adapt, but the market shifted in ways nobody could imagine.
What I'm reading 📚
AI is finding bugs faster than humans can fix them: How enterprise security teams must adapt
~By Steven Vaughan-Nichols, Zdnet
AI labs can't keep their stars
~By Ina Fried and Madison Mills, Axios
Cloud giants pour nearly $600B into capex as AI demand surges
~By Dan Robinson, The Register
What I'm watching 📺
The double-debt problem
~Robert Rose, Content Marketing Institute
Look who's talking 👄
"AI has handed people like me an unfair advantage. I got 40 years of game tape in my body doing this badly, then a little less badly, then a little less badly than that, and then sometimes I do it well. And that accumulation is exactly what these AI tools can multiply. When I use AI to build a strategy, I can feel what's strong and what's hollow and what needs to be pressure tested."
~Robert Rose, Content Marketing Institute from The Double-debt problem video.
