Tech talk comes to standstill once you utter AI. Henceforth, all discussions revolve around AI. All that is being in this frontier area is limited to very few companies, some being well established IT behemoths and others being promising startups, which have been showing promise for quite some time now, the most outstanding example being OpenAI. Artificial General Intelligence, AGI, seems to be the final super prized accomplishment, but that will happen only through AI research and product creation’s ongoing battle.
What AI will magically transform and who will his job and new jobs come into existence can be nagging questions in sometime from now, what is critical right now is the shortage of AI talent, more so the proven one. Only they help any company or AI lab to be in the top in the topsy turvy AI journey. Time, talent, proven capability, and delivery are all essence. Time is running out. All might not be fair in love, war and politics too, but it’s fair to give any hike, equity and perks to the deserving AI talent. A recent headline blares. “How Anthropic is snatching top talent from OpenAI and DeepMind.”
The AI industry is right in the middle of a cut throat talent war, both tech giants and startups are competing fiercely for top researchers, engineers and executives. If we compare retention rates across all AI labs, then Anthropic emerges as the leader. Nearly 80% of people who joined two years ago are still there. DeepMind is not far behind at 78%, while OpenAI’s retention is at 67% and Meta at 64%. Anthropic is preparing to launch its first employee share buyback program. To retain talent you have to go beyond solely relying on high compensation packages.
No one wants to talk about it but it’s the compensation package that makes the difference. One million or more salary package offers are very common for specialised skills. Some of the top names in AI are making the jump; Sholto Douglas, Niki Parmar, Neil Houlsby, Pavel Izmailov, Jan Leike etc. There is a shortage of talent. Knowledge required to train models at scale is really not taught in schools. It is possessed by few individuals who have experience in it. The global AI talent pool is limited, with only a few thousand researchers capable of developing models such as ChatGPT or Gemini. Non-compete agreements of one year and six months are becoming the order of the day. In a bid to retain, DeepMind is shelling out annual compensation packages of up to $20 million.
WHILE OTHERS MIGHT LOSE THEIR JOBS IN THE DAYS TO COME, AI EXPERTISE IS HAVING A WINDFALL.
Sanjay Sahay
Have a nice evening.