My blog

Tag: AI

  • Resetting the baseline

    The current AI slop flooding our timeline is setting a new baseline.

    The previous baseline was people with no access to creative sources posting text only or badly done graphics. Now, anyone can add a prompt and generate an image, but… they ALL look the same and that is the new baseline – media drowning in a sea of sameness.

    I find myself rapidly scrolling past most AI driven ads/media with a subtle ‘ew’ resonating in my being, but stopping at the content that appears thoughtful and human made.

    Creative work is not dead, the assumption that it can be replaced with an algorithm will eventually show that you do need creativity.

    There’s still lots of space for creativity, it’s just shifting and doing bit of shape changing, the creativity of the future is grounded in authenticity and craftsmanship.

  • Let’s fully trust the algorithm shall we?

    How many times has Facebooks ‘people you may know’ actually been correct?

    Out of the 70 plus people it feeds you daily, how many of those do you really know? Some are old acquaintances that you don’t want to reconnect with, some are there because someone you knew has had their mobile number recycled so you clearly DO NOT know that person? How many of those people do you really want in your intimate circle?

    This is why human judgment is crucial in the age of AI. It’s trained on past human experiences/knowledge, responding to the present in the absence of the subtleties that make us human.

    I see some are abandoning creative input in the favour of data, idolising the cold hard facts. People try a bit too hard to quantity, commodify and replicate creativity through its polar opposite: numbers, patterns and data sets.

  • The future of the new work

    There are two basic routes some companies seem to be following when restructuring with AI.

    • Using skilled subject experts to teach their AI models and then replacing them with less skilled (lower paid) staff to stay and use the AI and training them as they would need to increase their skillset
    • Using less skilled/outsourced services to teach their AI models and employing multi-skilled more experienced staff to stay and use the AI

    Neither of them are great ideas as you need quality input as well as quality AI ‘Operators’. The focus is on short term financial wins and the fact that they’ve been overpromised what AI can do.