Ouch, testing and quality are even more important these days, in the age of #AI Unfortunately, the market wanted them to release too early… before it’s fully tested and ready…and that same market just punished them 🤡 #aitesting #testingai #softwaretesting https://lnkd.in/gPiM4iDP
I think this answers the question as to why testers should have a high pay.
Jason Arbon Is this a story of commercial pressures to release the product versus product readiness? I don't wish to be insensitive, but were there similar factors at play which led to the Challenger Space Shuttle disaster?
Looks like there is conditional signoff given 😂
The concept of testing is very different for AI than traditional software. It is much more difficult to judge whether an AI service is "ready" without actually rolling it out to the public. When ChatGPT was released, the team made it very clear upfront that it's experimental. But the bar for Google now is so high even the result of a first single query made wave.
In my opinion ChatGPT only delivers answers similar to the status quo of top search engine results, though with regards to details, such as analysis between x and y it is extremely 'intelligent'. Example 1: ChatGPT confused a ssh public key as being private. On a quick search on Google, the top results were misleading, there was a post which contained mistakes, making that confusion too, a public key was presented to be private in a blog-like post, and that information was being propagated by other websites. Example 2: When asking chatGPT what is a JMeter agent, it told me that a JMeter agent is one of the remote running instances. But actually there exists the notion of server agent, which is totally different, and delivers metrics. When searching on Google, the top result was saying JMeter is a remote execution agent, which is true, (but that is different than server agent). Considering this matter, it would be worth checking if the bad answers of Google AI were either worse or better than chatGPT's. Share value is fluctuating. I think it is normal they go up and down some times, though I don't follow those price charts.
That hurts! Whether it's AI or whatever, this again reminds us of the importance of testing. TEST the damn thing before you bring it to the market.
The true fact about AI: it's Artificial.
Miss in QA probably costed $110 billion 8% of googles market cap. A good learning and step back for all fast paced Startups who are highly focused on market first and market catch-up, and valuations, which are Ofcourse very important but need to understand your product or software is as good as it’s quality. Please understand quality is about customers. The way you cannot ignore customers you cannot ignore quality. For the sake of race. 🙏
Helping to make the AIs a little faster/efficient!
1yDisagree with the premise. The market punished Google since Google has positioned itself as the leader in AI for years. However a startup released a sector defining product and now they are scrambling to catch up. This is a much higher level problem then "testing".