The downsides are a dependency on Internet access and a potential lack of privacy. Internet access is almost ubiquitous,

Privacy is not an issue with Apple. They too can search photos using machine learning, but they do it on your device. Apple has no clue about what’s in your photos, and because they make their money selling hardware, they don’t care.

http://old.cocir.org/media/cla/video-Oxford-Doncaster-v-en-gb-1rdr-6.php

http://startup.munich.es/dyn/video-Valenciennes-Nancy-v-en-gb-1rcb30122020-28.php

http://www.ectp.org/kzz/videos-Raja-Club-Athletic-Teungueth-v-en-gb-whi-.php

http://team.vidrio.org/rtr/v-ideos-Rodez-AF-Chamois-Niortais-v-en-gb-1cxi-18.php

http://molos.bodasturias.com/qxz/Video-Linares-Deportivo-Sevilla-v-en-gb-1ozy-19.php

http://old.cocir.org/media/cla/video-Oxford-Doncaster-v-en-gb-1scd30122020-13.php

http://startup.munich.es/dyn/Video-Valenciennes-Nancy-v-en-gb-1bpv30122020-11.php

http://elta.actiup.com/mto/v-ideos-Raja-Club-Athletic-Teungueth-v-en-gb-qyp-.php

http://www.ectp.org/kzz/Video-zamora-v-villarreal-v-es-es-1syl-10.php

http://molos.bodasturias.com/qxz/video-Doxa-Katokopias-APOEL-Nicosia-v-en-gb-1gfn30122020-.php

http://startup.munich.es/dyn/v-ideos-Sochaux-Dunkerque-v-en-gb-1gqa-.php

http://old.cocir.org/media/cla/Video-Tottenham-Brentford-v-en-gb-1uqt-.php

http://molos.bodasturias.com/qxz/Video-Doxa-Katokopias-APOEL-Nicosia-v-en-gb-1uwe-9.php

http://team.vidrio.org/rtr/Video-Rodez-AF-Chamois-Niortais-v-en-gb-1owj30122020-26.php

http://startup.munich.es/dyn/videos-Sochaux-Dunkerque-v-en-gb-1qat30122020-4.php

http://old.cocir.org/media/cla/v-ideos-Tottenham-Brentford-v-en-gb-1ufx-14.php

http://www.ectp.org/kzz/Video-zamora-v-villarreal-v-es-es-1rcp-25.php

http://elta.actiup.com/mto/videos-Raja-Club-Athletic-Teungueth-v-en-gb-ikd-.php

http://startup.munich.es/dyn/video-Sochaux-Dunkerque-v-en-gb-1awa-24.php

http://old.cocir.org/media/cla/videos-Tottenham-Brentford-v-en-gb-1crp30122020-21.php

e you would put any relevant experience to the data scientist position, but it was not a permanent/internship. It could be an organization experience, event experience, part-time work, and anything else you deemed important for the position.

So, how was I going to create 105,000 JPEG images? Adobe Lightroom Classic, as it is known now, is superb at doing that sort of thing. Getty uses JPEG images rather than RAW files, so that’s always been part of my workflow.

I ended up wasting 5 hours because I stupidly put all the images in one folder for convenience, hoping to upload all the files in one go. Repeatedly returning to the iMac to upload collections of images seemed tedious.

When I told Apple Photos to import the images, it seemed to start well, so I left it to get on with the import. I returned a few hours later, and the iMac seemed to have hung.

Apple offers storage packages up to 2 TB, depending on where you live (bizarrely). Two terabytes costs £6.99 per month in the UK, and I was already subscribing. That clearly wasn’t going to be enough.

What problem was I trying to solve? I wanted to search all of my images with AI, but you don’t need full resolution images for that. I’d convert mainly RAW images into HD JPEG ‘snapshots’.

Unfortunately, Apple Photos doesn’t seem to work like that. It seemed to have copied the images off the external SSD on to the main iMac hard disk. It then seemed to attempt to swap photos with iCloud. This might work well with a few images, but not with 100k images.

What does have more impact is missing out on ever-evolving machine learning, honed constantly by a global AI mind. Apple does a major iOS update annually, so you have to wait to get the latest and greatest. AI is far from perfect, so the prospect of frequent timely improvements is almost a necessity.

I behind to wonder about disk space. My JPEGs were on an external SSD, but I was importing them into Apple Photos. Apple Photos is supposed to scale images to suit available disk space.

I went to Lightroom’s library module and selected all the images. I activated my export pre-set and choose a SSD folder as the destination. Five hours later, Lightroom had exported 340 GB of images.

The downside to a local device search is that it’s local. At the time of writing, Apple does not share its analysis of your photos between your devices. Every device has to reanalyse your photos, so there will be a short delay before you can search your photos on a new device. In practice, that doesn’t have much impact.

That should take up much less space than 4 TB. I’d upload the snapshots to Apple and maybe Google. Apart reducing the storage required, my upload times would be substantially reduced.

I’d use the snapshots to find images, and then I’d jump over to the full resolution images on my SSDs. Would this work on my iPad Pro? Yes. I connected a SSD using a USB-C adaptor, and the drive appeared on Files.

Google offers 15 GB of storage for free, so I had to look at paid options with Google One. The cheapest plan that offered enough to storage was the 10 TB plan at $99.99 per month. This was way more than I wanted to pay, and it didn’t seem to be available in the UK either.