04.18.2025

ForwardThinking 🤔

IT fundamentals Still apply — even in AI era

When you have been around enterprise computing long enough, you start to see history constantly repeating itself. With each new technological wave, it feels like we are being told that the old rules don’t apply this time. AI agents are the latest hot trend, but guess what: fundamentals always apply, regardless of the technology.

There are no shortcuts, my friends, and that’s even more true if you are part of a regulated industry. As my colleague Ed Sim pointed out in a LinkedIn post this week, agents are everywhere, but companies that really want to stand out from the crowd are going to have to solve what he called the last mile problems, which turn out to be all those things that large organizations care about: “security, privacy, compliance, integrations, determinism and repeatability, accuracy, memory, cost, orchestrating swarms of agents, and even more security.”

Photo by Peter Conrad on Unsplash

As I pointed out in this space last week, it was a little mind-boggling to me that we have been creating agents without communications protocols up until recently, yet that’s just one step of many. Sure, they have to communicate with each other – obviously – but they also need to have all of these characteristics that Sim listed. Without them, you can’t hope to build and implement complex, enterprise-ready agentic solutions that large organizations are going to be lining up to buy.

I feel like we’ve been here before

I’ve been thinking about the current agentic moment this week and how it reminds of another time, not all that long ago actually. Think back just four years to 2021 when RPA, workflow and low code tools were coming together to have a moment of their own. UiPath, the RPA leader, was riding high, ready to go public and sporting a $35 billion valuation. No code and low code tools were proliferating and workflow automation was getting oodles of attention. There was a ton of M&A in these areas too, as the larger players like SAP and ServiceNow jockeyed for position. (Sound familiar?)

Illustration of man with a laptop working with a robot on a blue background.

Illustration by Mohamed Hassan on Unsplash

But it didn’t really work out the way people expected. While RPA was effective for automating repetitive, rules-based tasks, it was rigid and had trouble scaling and adapting to changes in the workflow, something that agents are supposed to excel at. Even UiPath founder and CEO Daniel Dines acknowledged in a recent interview with The Verge, that agentic AI and workflow automation are coming together in a new way.

“The way we are seeing the adoption of combined agentic AI and automation is by putting a workflow technology on top of it,” he said. It’s worth noting that UiPath’s stock price has plunged since it went public in April, 2021 from the opening price of $56 a share to $10.45 with a market cap of around $5 billion as of this week, well below that lofty pre-IPO valuation.

A lot of trends we saw coming together in 2021 are suddenly back in vogue, but this time AI agents are displacing everything from coding to workflow to process automation.

Ignore the noise

It’s easy to succumb to the hype, but Thomas Squeo, Americas CTO at tech consultancy ThoughtWorks, says fundamentals still apply with agents. “At the end of the day, you have to manage it as a piece of enterprise software because vibe coding alone (coding with the help of AI agents) is only going to get you in trouble,” he said. Even with agents in place, maybe especially with them in place, large companies still need to pay close attention to how they manage risk.

“I think that the biggest issue that I've seen is that enterprise risk is the force of gravity that everybody has to pay attention to,” he said. “That’s because as soon as you have to disclose how you're maintaining a production system, and what the risk profile of that system is in the enterprise, especially for publicly-held companies, that disclosure risk makes everybody more conservative.”

"I think that the biggest issue that I've seen is that enterprise risk is the force of gravity that everybody has to pay attention to."

The problem with hype is that it puts pressure on organizations to implement technologies they might not be fully ready for. Yes, you need to be exploring how agents could potentially transform your organization, but ultimately, there are no implementation shortcuts for enterprise companies, no matter how loud the buzz gets. Your company still has to adhere to the same fundamental principles of enterprise computing, whether it’s agents or any other shiny new technology, and there’s just no getting around that.

~Ron

What's new on the blog 📰

How Akamai’s CIO saved $100 million annually by bringing public cloud workloads back on prem

Akamai CIO Kate Prouty has managed to save oodles of cash for her company by questioning every public cloud workload. By moving many workloads back on prem, she found she could control costs better.

"There was a huge opportunity here for us to rethink the way we leverage third-party cloud services," she said.

Read the full article >>

Analysts identify key takeaways from Google Cloud Next 2025

Even though I couldn't attend Google Cloud Next last week in Las Vegas in person (and had some serious FOMO), I asked three top analysts to identify the biggest announcements. Here's what I found:

Read the full article >>

Shot from above of full stadium at Google Cloud Next 2025. There are two huge screens with the word cloud on them with rainbow colored light beams behind the words.

Photo courtesy of Google Cloud.

This former Salesforce exec is using AI and data to transform higher education

Phil Komarny, a former Salesforce executive, is trying to transform higher education and the college experience using AI and data. His goal is to help students get the most out of their time at college, and beyond, throughout their working lives.

"My thesis is that if there is a better way to treat student data and give them ownership, while making it more accessible, why wouldn't we do that,” he said.

Read the full article >>

Andrew Ng warns AI hype is distorting its true value

Andrew Ng spoke via Zoom at the CrewAI Agent Summit earlier this month. One thing that stuck out was how he pushed back against self-serving PR hype coming from the biggest vendors, causing unreasonable expectations and fear, and masking the incredible things AI can actually do today.

"The PR hype is disconnected from the business reality of progress. We frankly [need to] ignore most of the hype and just keep doing the work that creates value," he said. (Note that both boldstart an Ng are investors in Crew.)

Read the full article >>

Andrew Ng speaking at TechCrunch Disrupt in San Francisco in 2017.

Photo courtesy of TechCrunch on Flickr. Used under CC BY 2.0 license.

News of the Week 📣

Figma Files intentions to go public (eventually)

In September, 2022 Adobe dropped a bombshell when it announced it intended to acquire Figma for $20 billion, a company many considered a rival. The deal eventually fell through under the weight of regulatory scrutiny, but this week the company filed a confidential S-1, signaling its intentions to go public.

The confidential part means that it has filed S-1 paperwork, but doesn't have to reveal the contents publicly just yet. It can also delay the date basically for as long as it wants to, but when it does reveal the timing, it will also have to let us see the financials.

Figma CEO Dylan Field at Figma Config conference in 2024 wearing a 'Now with AI' t-shirt.

Figme CEO Dylan Field. Photo courtesy of Figma

Figma, a cloud-based, collaborative design platform that's extremely popular with designers, seems to have managed to thrive in spite of the failed merger. The $1 billion kill fee had to help a little, but the company has continued to grow along the way.

Last July, it reportedly had a valuation of $12.5 billion, well below Adobe's generous offer, but still higher than their $10 billion valuation in 2021. And we all know that 2021 valuations were often a tad generous. Today, the company has 1600 employees, and CNBC reported it earned $600 million in revenue last year.

When this goes, it could be a major IPO—especially given that IPO activity seemed to be gaining momentum earlier this year, only to stall again amid the stock market turmoil related to the on-again, off-again tariffs.

Activist investors Elliott Management take $1.5B stake in HPE

This week, news emerged that activist investors Elliott Management, have taken a $1.5 billion stake in Hewlett Packard Enterprise. Elliott, known for targeting companies it views as undervalued, often pushes for operational cost cutting measures and reorganizing, which typically results in layoffs.

Reports suggest they actually want to make one big employee change, replacing CEO Antonio Neri, who has been with the company for decades. Neri joined HP in 1995, and moved over to HPE when the company was split into two companies in 2015. He was promoted to CEO in 2018.

Photo courtesy of HPE

HPE makes servers that get installed in data centers. You may have noticed this year -- we certainly have -- that there is a lot of data center building activity going on in 2025, a trend you think would favor HPE, but the stock is down 30% for the year, leaving it vulnerable to activists like Elliott.

One thing the company was counting on to increase revenue was being able to push through the $14 billion Juniper acquisition announced last year, but the deal has run into obstacles with the Department of Justice filing suit earlier this year to put a stop to the merger. It's worth noting that analysts I spoke to expressed skepticism regarding the suit, but it's not doing anything to improve the stock price and Elliott is apparently looking for a change in direction.

Another week, another flurry of openAi news

What kind of week would it be if we didn't have OpenAI news to share with you. The company keeps the news coming, that's for sure, whether it's scuffling with former employees, a new model (but of course), and an acquisition rumor.

Let's start with disgruntled former employees. The group filed a friend of the court (amicus) brief this week, joining Elon Musk's lawsuit opposing OpenAI's proposed conversion from non-profit to for-profit organization, and the axing of the "for the benefit of humanity" part of the original charter. One of the signers went so far as to call OpenAI CEO Sam Altman, "a person of low integrity." Ouch.

Altman didn't help his case when TechCrunch reported this week that the company has altered its safety protocols in a way that seemed to suggest that if competitors loosened their safety measures, OpenAI would follow suit.

Sam Altman appearing at TechCrunch Disrupt in San Francisco in 2019.

Photo courtesy of TechCrunch on Flickr. Used under Creative Commons Attribution 2.0 license.

Meanwhile there was a new model, GPT 4.1, which the company says is aimed at improving AI coding. It may do that, but HubSpot CTO Dharmesh Shah, who is usually an AI cheerleader, damned the model with faint praise calling it "pretty good" in a LinkedIn post, and even served up a few complaints.

That could explain why rumors were flying this week that OpenAI was in talks to acquire AI coding tool, Windsurf with word on the street pegging the price tag at a tidy $3 billion. As Axios pointed out, given OpenAI's market position, regulators could have something to say if this deal were to come to fruition.

As if that weren't enough, CNBC reported that OpenAI originally had its eye on Anysphere, the creators of Windsurf rival Cursor, the current darling of the vibe coding crowd. But that would have come at much steeper price as Anysphere was reportedly raising last month at a $10 billion valuation.

What we're reading 📚

Person reading book, but all you can see is their hands holding the book. They are sitting on a hardwood floor steeped in shadow, creating an interesting look.

LivPhoto by Blaz Photo on Unsplash

The New Stack

Vibing Dangerously: The Hidden Risks of AI-Generated Code

Darryl K. Taft

Palo Alto Online

Silicon Valley crosswalk buttons apparently hacked to imitate Musk, Zuckerberg voices

Zoe Morgan

Axios

AI's climate impact is still a black box

Megan Morrone

The Verge

Wikipedia is giving AI developers its data to fend off bot scrapers

Jess Weatherbed

Look who's talking 👄

"No single country in the world can ever do everything great: a rising middle class does not want to sit in a hot, dirty, dangerous coal mine or manufacturing facility to earn the same or less they could make in a service job. They prefer to code a social-media app while sipping matcha smoothies and playing foosball. "

- Esteban Kolsky, An (Alternative) Method to the Tariff Madness

Please subscribe to the newsletter and feel free to share.

3350 Virginia St, Suite 262, Miami, FL 33133

ForwardThinking 🤔

IT fundamentals Still apply — even in AI era

I feel like we’ve been here before

Ignore the noise

What's new on the blog 📰

How Akamai’s CIO saved $100 million annually by bringing public cloud workloads back on prem

Analysts identify key takeaways from Google Cloud Next 2025

This former Salesforce exec is using AI and data to transform higher education

Andrew Ng warns AI hype is distorting its true value

News of the Week 📣

Figma Files intentions to go public (eventually)

Activist investors Elliott Management take $1.5B stake in HPE

Another week, another flurry of openAi news

What we're reading 📚

The New Stack

Vibing Dangerously: The Hidden Risks of AI-Generated Code

Palo Alto Online

Silicon Valley crosswalk buttons apparently hacked to imitate Musk, Zuckerberg voices

Axios

AI's climate impact is still a black box

The Verge

Wikipedia is giving AI developers its data to fend off bot scrapers

Look who's talking 👄

Subscribe

FastForward #18: IT fundamentals still apply, even in AI era

You may also like...