☔ AI's Summer Recap: From heavy rain to Neural Networks. 🧠
Unmasking the Tech Behind ChatGPT, Future Workforce Predictions, and the Surprising Ways AI is Shaping Our World. Dive in for a Summer's Worth of Insights.
👋 👋 to all the 502 new FullStack HR readers joining during the summer!
If you aren’t subscribed yet, join the curious and like-minded leaders and HR folks by subscribing here:
Summer has flown by, best captured by the GIF below…
But still, taking time off during the summer is the A-game in Sweden. Almost everyone here takes at least four weeks of vacation, which is truly wonderful!
My family and I visited Åre, where my sister got married (yey!), spent time in Småland, and cherished moments of relaxation, thoroughly enjoying our time together.
While the weather might not have been on our side, there's an undeniable beauty in allowing oneself the luxury of downtime, letting thoughts drift without agenda.
For the initial weeks, I consciously steered clear of anything AI-related. However, AI's allure proved irresistible, and it soon reclaimed my attention, becoming my primary focus in the concluding weeks of vacation.
Because I'm on the brink of unveiling…something, you'll be among the first to know when it's ready.
But let’s say this if you have friends that are keen on learning more about AI and HR - invite them to subscribe to the newsletter - fun times ahead!
But for now, let's transition from summer reveries to the energy of fall.
And we’ll start the semester with a run-down on good AI stuff I stumbled across this summer.
Demystifying AI and LLMs.
Want to really understand AI? Spotify's Gustav Söderström explains how systems like ChatGPT work conceptually in an engaging talk. He covers the core techniques for natural language, image generation, and music composition.
While complex under the hood, Söderström argues the principles of neural networks and word vectors are intellectually intuitive. He uses visuals and analogies to explain how AI models capture statistics and "learn" representations. The talk aims to demystify AI, conveying the basic theory in an accessible way.
So if you want to debunk the hype and complexity around AI, Söderström's presentation is a great starting point. You'll come away with a solid grasp of how systems like ChatGPT function at a high level. A must-watch!
Future of Work and AI
Up to 30% of work activities could be automated by 2030, with lower-wage and less educated workers facing high disruption risk. Nearly 1 in 5 workers are already in highly exposed roles.
Sectors like healthcare, STEM, and management are poised for job growth, while office support, food service, and customer service may decline.
8.6 million Americans switched occupations from 2019-22, a 50% increase showing how quickly changes unfold.
Reskilling initiatives must be massively scaled up through public-private collaboration to support transitions.
Training should focus on social-emotional, digital, and higher cognitive skills that are harder to automate. Short modular programs can supplement more traditional training.
Organizations must shift recruiting to assess potential and transferable skills rather than just credentials. Hiring for aptitude and willingness to learn expands talent pipelines.
As automation advances, jobs can be redesigned to emphasize relationship-building, creative, and strategic work. Employees should be involved in envisioning how AI can enhance productivity rather than replace jobs.
The research makes clear, proactive workforce investment and a focus on constructive transition will be critical in the years ahead.
28% use ChatGPT at work.
Around 28% of workers use ChatGPT at work despite limited approval. 10% of surveyed workers say their bosses explicitly banned external AI tools, while about 25% (!) do not know if their company permitted the use of AI technology.
This is all according to a poll by Reuters/Ipsos that surveyed 2,625 adults across the United States and had a credibility interval of about two percentage points, a measure of precision.
There’s certainly room for us in HR to bring clarity (and training) to our employees here!
Bard (Google equivalent to ChatGPT) is now available in 230 countries worldwide. Free to use and fairly good - give it a try! (Or finally, ditch the VPN…)
Microsoft Co-pilot seems to be well on their path to launching widely later this year.
Googles Duet AI seems to be heading in the same direction.
The hunt for AI talent is on - A few weeks ago, Netflix advertised a position offering total compensation of up to $900,000 for AI professionals. Other big players, including Hinge, Walmart, Goldman Sachs, and Amazon, are hunting for AI specialists, offering salaries of $250,000 and above. On average, engineering positions at Prompt command a total compensation of around $130,000.
elevenlabs recently introduced input streaming, enabling users to seamlessly receive LLM responses and produce speech instantly – all achieved with an impressive sub-second latency. Now, when will candidates utilize this to answer a recruiter’s questions?
And don’t be a bot…
AI is no longer the new kid in town.
The hype around AI is cooling down, and we start to see people airing opposing views, e.g., that it’s only hype and what’s the real use case?
I have a long piece on this next week.