GPT-4: End of Graphic Designers? | Hacker News Reader

GPT-4: End of Graphic Designers?

I remember the same words being used for developers when ChatGPT launched, when Devin launched, when Cursor launched. What happened then? Instead of the end of developers, developers became super powerful with the help of AI tools. Who would have imagined we could speed-run a whole application in a single weekend with the help of AI tools? So, whatever happened to developers after GPT-4 and Claude Sonnet will happen to graphic designers as well. Tools like Cursor will come for graphic designers, and using that tool, they will become super powerful.

The interesting thing to watch in the coming months (weeks?) is who will win the "Cursor for Designers" title. There are a lot of big players in the market, like Figma and Adobe. Will the big players be the kings? Or, like Cursor, will a new player come and dominate the big players?

What do you think?

View on HN

I would counter that LLMs have made worse devs and they were pretty bad before. Being more productive and writing more code hasn’t made software better, it’s the contrary. If we’re more productive in producing code churn and could rely on the smarts of an LLM, shouldn’t we be moving back to the days of resourceful code and languages? There is no longer the excuse that X is hard, memory management is hard, pointers are hard, etc. And dare I say it, the death of multi-platform web apps. If a single dev can do the job of 3, they can maintain the cost and burden of 3 platforms as the LLM is doing the heavy lifting, right? We’ll save on energy (or balance it out as LLM farms use a lot), we’ll save time (faster code), we’ll produce less e-waste. These are real problems great devs think about.

The same will be true for artists. Terrible artists will have the perception of being great but will fall flat with bad artistic style and vision.

Yes. Good will shine out. Bad will be gone away.

I think your observation is not a good assessment of the impact that LLMs have on industries.

It's true that "developers became super powerful", but that also means that suddenly companies need less manpower to achieve the same business goals. This increase in efficiency means less jobs, because now companies suddenly need less people to handle the same workload.

Also relevant, a company no longer needs that many experienced developers because you no longer need to have in the payroll someone who did something before. There's now less arguments to hire a consultant to help a company with a new framework or tech stack, because a LLM allows devs to turbocharge their own onboarding. This means less jobs, and being able to have inexperienced devs handle tasks previously reserved for mid/senior developers.

I don't think any llm-based ai is going to displace good and great designers. If someone is wearing the badge of "designer" and all they do is crank out mediocre Photoshop edits, then they may have to worry, much like the people who couldn't code had to. These tools make people who don't understand the domain super confident in themselves, but they eventually have to learn somehow.

Reply on news.ycombinator.com