Simon Fletcher

Building the new era software development lifecycle @glideapps, previously @github

[Simon Fletcher @simonfl3tcher]

@mfishbein @alighodsi @a16z Knowledge and context extraction is now the hard part of agent efficacy. Whoever solves this will have a huge advantage over their competition.

[Simon Fletcher @simonfl3tcher]

Knowledge and context extraction from the business and humans continues to be a huge bottleneck for agent efficacy. We are working relentlessly @glideapps to solve this so we can solve it for ourselves and our customers

[Simon Fletcher @simonfl3tcher]

I haven’t written code in 18 months. I haven’t reviewed code for about 8 months.

I look back on reading/writing code with nostalgia. However, for me it was always a means to an end. It was the making of something that attracted me to software.

The future is bright for software engineers.

[DHH @dhh]

Software Maker > Software Writer

[Simon Fletcher @simonfl3tcher]

@omkarships No. Our job role will simply evolve. We just don't know where it will land yet which is causing the anxiety.

[Simon Fletcher @simonfl3tcher]

Unpopular opinion:

It increasingly feels like agent capability is becoming less of a bottleneck for applied AI at scale.

The harder part is everything around it: security, governance, and extracting knowledge from workers so that AI is effective

Ofc there is still a category of problems that need deep capability. But for the average everyday business task, I don't think more capability is the constraint.

[Simon Fletcher @simonfl3tcher]

@GergelyOrosz This is very similar to how our factory @glideapps
is set up. We have more boxes in after the production step, around runtime errors, bugs, and code quality, but in general, very aligned with this flow.

[Simon Fletcher @simonfl3tcher]

@levie Hard agree. My sense is that we don't need more capable models right now to harness AI's power at scale. We need the long and short-horizon workflows that enable organizations to scale AI effectively. More capable models generate more complex work that will amplify the need

[Simon Fletcher @simonfl3tcher]

@sweet_lil_adder @owenthcarey I will say though that the industry hasn’t caught up yet in terms of how to interview / hire in this new world. It will catch up. But systems thinking is still a valuable asset to have when entering engineering.

[Simon Fletcher @simonfl3tcher]

@owenthcarey I would say though that the industry hasn’t caught up in terms of how they need to interview / hire. It will but I don’t know when.

[Simon Fletcher @simonfl3tcher]

@owenthcarey Learn systems thinking over code level concerns

[Simon Fletcher @simonfl3tcher]

Customer obsession. It's how you win in this era.

[Simon Fletcher @simonfl3tcher]

@KomalWCode Systems thinking

[Simon Fletcher @simonfl3tcher]

@prasenx AI cannot generate good code or architecture without human steering.

In my experience, if you lean in and give it the guardrails and backpressure with your taste and judgment encoded, it can do a pretty good job.

[Simon Fletcher @simonfl3tcher]

Once human feedback diminishes on that sample size, you go fully autonomous for that size. Then you increase the sample size and repeat.

[Simon Fletcher @simonfl3tcher]

One of the main challenges of building agentic workflows & loops is getting feedback from humans. LLMs are only as good as the context you give it. Without the necessary context its ability to have a shot on target is drastically diminished.

[Simon Fletcher @simonfl3tcher]

In order to recognize the value of AI at scale you have to start small (knowing how much you are leaving on the table) and the output needs to be super high signal. The loop then needs to fold the feedback back into the loop so that next time the agent has a closer shot on target.

[Simon Fletcher @simonfl3tcher]

The problem with building long-horizon agentic loops that matter is that the time to feedback is so long. It's a slog. So much so that our second generation of loops is coming with built-in evals and dry-run flow

[Simon Fletcher @simonfl3tcher]

Once the platform is reliable enough for long-horizon work, the role becomes delivery, verification, and customer obsession. That's a product role.

I believe that the software engineers who thrive will pick a lane and go deep.

[Simon Fletcher @simonfl3tcher]

The software engineering role as we know it is atrophying. I see two paths to replace it.

Software engineers will either:

1. Move into platform engineering to enable the rest of the organization.
2. Into a product role.

[Simon Fletcher @simonfl3tcher]

At @glideapps, bugs, security, code health, reliability, and architecture of our platform are all handled by multi-agent, long-horizon loops with minimal human in the loop. The traditional engineering toil is being abstracted away from engineers, allowing them to focus on product solutions to a stated problem. Here lies the rub. This is not an engineering role. This is what product managers have been doing for years.