Artificial intelligence is all over the place! Google is working on an AI model that can translate hard-to-read handwriting – meaning you won’t have to second-guess your doctor’s prescription the next time you head to the pharmacy. The Q-bear uses sophisticated AI technology to “translate” a baby’s cries within 10 seconds, so parents can deliver the milk/snuggles/diaper change more immediately. Meanwhile, researchers in Copenhagen, Zurich, and Paris have created technology that allows them to interpret oinks and snorts into whatever Piglet might be feeling! As you might have expected, translations and copywriting aren’t exempt from the AI wave. In fact, we weigh 5 different AI tools and how they could possibly make translators’ lives easier. Read on to check up on Copysmith, Perci – Britton Upchurch, CopyMonkey, Jasper, and Powerlisting.
GPT stands for Generative Pre-trained Transformer, and the 3 stands for 3rd generation. GPT-3 is a machine learning model created with open AI and neural networks. You might see GPT in this blog; now you know what it stands for.
The YLT team tested the tools to see if they would make work easier or not. They used the tools in Spanish and German to gauge ease of use and actual output. Copysmith, CopyMonkey, Powerlifting, and Perci handle Amazon listings, while Jasper is adept at managing general content.
How Do These Tools Help with Content Creation?
The 5 tools were better as “listing generators” instead of “listing translators.”
To quote Jana:
Generating copy for a certain topic was okay, but if you were looking for creativity and attention-seeking content for your social media channels, this type of content would not get you insane amounts of views or shares. Content is OK, but still nowhere close to the written word and creativity of an actual copywriter. Score: 2.5/5.
As it currently stands, the content creation industry is huge. According to Future Market Insights, the market was expected to be valued at $13.4bn at the end of 2022. Naturally, AI will go where the need is the greatest. Content creators have been using AI to generate ideas to propel their efforts forward. According to Forbes, content creators rely on AI tools for idea generation, creative content ideation and copywriting, text optimization for SEO, content ideas, finding influencers for their work, and even creating their own AI influencers, whose every move is controlled and managed.
Nevertheless, it looks like the machines won’t be taking over content creation just yet. The human component still wins; the AI tools don’t work unless there’s real human ingenuity pointing it in the direction it should go.
Where will AI in content creation go? At the rate it’s being developed, we’ll likely see progress in that aspect very soon!
How Well do AI Tools Handle Translation Work?
The biggest problem with these tools is that they didn’t match the original text when translating.
Here’s what Jana had to say:
Sellers always want to have all the main features of their product translated to other languages as well and this is something these tools still can’t offer. We could draw the same conclusion with Jasper, (even though not an Amazon specific tool) – decent in copywriting but definitely not great for translations, as their language option is still in BETA. Score: 1/5
Listings are tough to do in any language. That’s why YLT Translations doesn’t just translate listings; we localize them. Native speakers handle the keyword research and copywriting. Instead of bringing an English text through machine translations, we take the features, attributes, and benefits, and craft a new listing inside the playing field of the target language. Many machine translation tools don’t do that, which is why expression may be inaccurate, keywords are often incorrect, and many features are left out.
Nevertheless, machine translations are growing rapidly. Take LanguageWire. The award-winning Danish tech-enabled language service provider is leading the pack with the greatest percentage of its revenue coming from Post-Edited Machine Translation services. Nevertheless, human beings still drive the technology. The tool’s Chief Technology Officer, Roeland Hofkens, said: “We have an in-house team of experts who have built our translation engine from the ground up, allowing us to focus on providing the highest level of translation quality and adapting content to the specific needs of our customers.”
Verbum, meanwhile, claims to provide 5 previously unavailable near-real time communication capabilities in one single tool:
- Automated verbal translation; it translates up to 82 languages and 40 dialects, so up to 60 participants can hear the conversation in their native language
- Multilingual closed captioning, so meeting participants can read auto-generated closed captioning in their native tongue
- On-the-Fly transcription, where transcripts can be auto-generated for up to 50 participants/languages
- Instantaneous translation at on-site events – Verbum can deliver the text of their speeches to the attendees’ mobile devices in their native languages
- Multilingual online chat, where customers can interact with brands and services in their native tongue, even if customer service speaks a different language
Let’s not discount Meta AI‘s research into speech-to-speech translation, which is indicative of the funding flowing into the development of sophisticated machine translation tools.
Nevertheless, none of these tools are ready for eCommerce. Machine translation is developing very quickly – let’s see if it’s ready for the Amazon listing in the next few years.
You Can’t Afford to Mess Up Your Keywords!
Keywords and Amazon SEO – the backbone of any successful listing. Instead of translating “barbecue gloves” into the target language, our translators actually generate new keyword research in the target language itself. Every language has its own way of calling a product, so it’s crucial to use the exact keywords in the target language. Otherwise you may miss out on what your target audience calls your product in their native tongue.
So, how do the 5 AI tools stack up with Amazon keyword research? Jana says:
Inserting keywords is little to non-existent, and the ones that are inserted, don’t often match the criteria of relevant keywords. Score: 2/5
Given the tools are ineffectual with listing copywriting, it stands to reason their SEO is similarly weak. Many SEO AI platforms generate keywords using data gleaned from thousands of online documents, from which they choose keywords to suggest to the content writer. If the copywriting – so, the foundation – is erroneous, then chances are the AI would have barked up the wrong tree to generate the keywords.
Even if AI isn’t ready for eCommerce SEO, it’s no secret that AI does very well at generating keywords on Google. AI can generate keywords from a massive amount of data – a robotic gargantuan task that humans wouldn’t enjoy much. Also, AI runs tests to find out which keywords perform the best. To quote insideBIGDATA’s report, “It’s a lot more effective to let an algorithm deal with another algorithm.”
The same report however posits that human SEO specialists bring a customer-centric approach to the work. AI’s vocabulary is limited to the data that it spans, so it tends to repeat itself every once in a while – thereby spamming your audience, something Google really doesn’t like to do. So, even on Google, keyword generation needs to be tempered by the human approach.
What’s the Conclusion?
Jana believes:
There will pass much more time before AI tools can take over human translation work, especially with keyword research.
I can definitely see this happening some time in the future, but not for another couple of years or so.
It’s just a matter of time before AI infiltrates eCommerce; so much capital goes into research in machine translations, copy generation, and SEO, that AI’s sophistication will soon be ready for eCommerce. In the meantime, humans and robots still work together. Algorithms do what algorithms do best, like culling massive amounts of data, while human beings add that, well, human quotient to the work.
Do you believe in AI’s capabilities? Did YOU download your Lensa AI avatars? Share them with us!