In 2015, I found myself talking to a smart entrepreneur who had just raised millions of dollars in venture capital from a prominent venture capital (VC) firm. He was flying high after getting substantial press and recruiting a number of early employees. His business also wholly depended on a popular open source project.

As part of his strategy, the CEO gave a core open source developer in the project equity. With this, he boasted that he would get critical information sooner, influence the roadmap for his company’s benefit, and recruit developers from the project to his company.¹

I was doubtful this connection would ever be publicized by the open source developer. Would he speak glowingly about the company at a conference, despite private reservations? How would he influence the decisions of his fellow developers? Would developers early in their career follow his missives without adequate reflection?²

Computer engineers regularly deal with exploits on their technical systems, where an external party takes advantage of a bug or vulnerability. An “exploit” or social engineering of our personal information sources — motivated parties influencing subreddits, conferences, meetups, blogs, journalists, professors, industry analysts — should be perceived similarly. As this example shows, even your peers can be influenced.

As engineers, it’s valuable to develop the antibodies that enable us to make the best decisions — both for our teams and for our own careers — despite the desires of these motivated parties. Beyond identifying ways we can be influenced, I list out three techniques to help us make better decisions.

The War to Influence Engineers

This is just one example in a larger war that rages in this global era of computing: that for the hearts of engineers.

Some of the many groups that will try to shape our views include:

• Developer tools teams who want to sell their technology or grow their platform
• Companies who wants to hire us
• Open source developers who want to recruit contributors and users
• Venture capitalists who want to fund us
• Training programs that want to sell their courses
• Consulting firms that want us to hire them
• Social networks that are trying to maximize engagement, readership, and revenue

The content they create and sponsor (press releases, meetups, blog posts, conferences, speaker talks) play a critical role in the decisions we make — from the technology we adopt to the companies we join.

A press release indicating large traction (often with many missing numbers) may be a primary reason to join a startup. A blog post or popular talk on the latest database technology — like NoSQL — may be a reason to experiment and later switch. A college lecturer may partly be there to pitch their own company’s technology. The most upvoted Hacker News (HN) posts over a year may influence what technologies we choose to adopt. And yet, as consumers of this information, many of us don’t always critically examine the information and its source.

In startups, bad technology choices made based on these inputs can divert the focus, putting the mission at risk. For an individual engineer, learning the wrong technology can waste our time. Joining the wrong company that sells us false promises can set our career back.

Even though I point out the issues in these case studies, I am far from a dispassionate scientist when on a mission for my team. Given that, it’s at least important to ensure that all engineers have the general tools to understand some of the messages we see. And in this era of fake news, selective facts, and troll armies, these lessons can also help us assess information far beyond engineering content.

Hidden Influences

Unlike previous generations of marketing, the external influences in engineering media today are often disguised.

The purpose of a press release or a sales call is clear. When this content is used to give a talk at a conference or inspire someone else to write a blog post, they blur the line. With a powerful marketing budget or fundraise, it becomes increasingly easy to publicize and convince others with one’s message, regardless of the veracity.⁵

For all these reasons and more, my most thoughtful friends use their own networks to understand reality instead: college/graduate school friends from select universities, frank talks with portfolio CEOs (for investors), ‘off the record’ conversations with connections at the various platform providers, trenchant blogs from front line engineers. Their information sources are less influenced by self-interest, providing a more objective view of the world. People can also say things privately they wouldn’t feel comfortable sharing broadly.

These friends of mine also have a deep-seated desire (specifically, a financial or technical motivation) to get at the truth, and so aim to understand the incentives driving every one of their sources. Though they may be outwardly agreeable, internally they’re deeply critical thinkers.

How Engineers Can Make Better Choices

These issues have a basic lesson: nothing is a replacement for surrounding oneself with primary sources, a community interested in the truth, and thoughtful engineers.

Beyond the self-interested factors, engineering media can be distorted for other reasons. For example, each upvote on Hacker News is equal — meaning that the world’s most thoughtful expert on a topic has the same voting power as someone inexperienced. Nearly every social media algorithm favors engagement over any other metric, meaning that what we read is simply a function of what those around us want to read (this echoes my research on what types of deaths are covered in a leading US newspaper).

For now, I’ll suggest three simple points that go far in assessing engineering media:

1. Identify incentives behind content that informs important decisions

First, consider a writer/speaker’s motivation — both directly (an employee or investor in the company) or indirectly (a public speaker who is a consultant looking for a gig or employee looking for a job/advisorship).

When I read a post or hear a speaker, I often ask myself a few questions:

• Background: What is their technical background and why are they well suited to speak on this topic?
• Objectives: What are their personal objectives in speaking/blogging? (e.g., recruiting, personal brand, new business); there almost always is one
• Conflicts: What conflicts might they have and how does this influence the message?
• Allies: Who else wants this content out there?
• Distributor Incentives; Why is the distributor (social network, media publication) putting this content in front of me? (who are the upvoters, what is the business model, etc.)

When we see a lot of press from a company, it’s often a sign that they want something from us. Their employees have consciously spent the time to sit down with a journalist or taken the time to write a post.⁹

2. Debunk claims

Second, inculcate a desire to debunk claims.

For example, companies regularly trumpet their (tailor-made) benchmarks that show their products in the best light. It’s often not that difficult to run our own benchmarks that reflect our real world use cases.

The direct reading of codebases is a powerful corrective (where available). It’s a primary source that doesn’t have ‘spin’ built in.

Looking back on trends in technology and where they ended up is another approach, which is akin to backtesting¹⁰: it’s easier to tune out hype, when we think about how people in a few years may perceive it, based on things we’ve seen in the past.

Say something in online communities when you disagree or see egregious incentive issues (here’s one favorite recent example on AV1 vs HEVC). On HN, this could be comments or a post of your own, even if it’s with a throwaway account. New social network designs might let us award karma to people who list out the key incentive issues in the future.

3. Build a community of respected peers and thoughtful experts

Third, examine the community that you lean on to determine important technical decisions.

It’s important to have a community of people who benefit from the truth, think critically, and have content expertise. This could be former employees, classmates, or your coworkers. You can augment this with the most thoughtful engineers you can find. Online, you can often find some of these groups in the comments of niche HN posts and open source mailing lists/chat rooms. In your early years, finding a thoughtful mentor is critical.

In many ways, this approach mirrors the web of trust in computer security, where we look to trusted friends — and people these trusted friends are connected to.

This point underlines the benefits of a thoughtful engineering curriculum taught by dispassionate teachers. I’m unaware of any school who teaches early engineers how they should best consume engineering media — or how to choose a technical tool. Often, this means that we’re then vulnerable for a few years, until we’ve incorporated the lessons from working on a good team.

Having a few thoughtful experts in our community is also critical when assessing deeply technical decisions like what database to use or how best to architect a system. The danger with some experts — the non-thoughtful variety — is that their confidence in their abilities means they don’t question many deep seated assumptions and can’t make adjustments when the world changes. Thoughtful experts have both content expertise and strong opinions that are weakly held.

Coda

We engineers share some similarities with the machine learning algorithms we design. In both, data is used to make inferences about future decisions. As such, it’s important to have a thoughtfully curated training data set — and to adjust for any errors when our engineering inputs are compromised.

I’ve hardly touched on all the other questions we should be discussing, beyond encouraging critical thinking and finding trusted networks of thoughtful engineers. For example, should we debate alternate Hacker News and Twitter algorithms with the hope of creating better engineers? Should there be conflict disclosures for open source developers? Should popular engineering blogs such as Free Code Camp have a board of engineers who technically assess content and the background of writers? Should we teach engineering media literacy in bootcamps and CS programs?

Still, I’m encouraged by the fact that critical thinking and testing hypothesis is such a natural part of how engineers and scientists approach the world. This mindset has a critical role to play far beyond the day to day engineering problems we aim to solve.