Chatty AI and Protein-Predicting Algorithm Defined the Year in Emerging Tech

They also built supercomputers designed to handle the growing influx of data needed for evermore complex AI applications. 

Emerging-tech startups kept feeding the innovation pipeline, though many were running on capital raised before mounting uncertainties in the second half of the year saw venture investors pulling back from funding deals.

Here are a few tech milestones of 2022, a year that many in the tech sector are likely happy to leave behind:

AI Generates Chatter

Years of R&D into generative AI capabilities came to fruition this year, highlighted by the release of ChatGPT, a free AI-powered chatbot with a humanlike ability to produce original, coherent text from a handful of prompts. ChatGPT, which made its public debut as a prototype in late November, and caught on quickly with millions of internet users, was developed by OpenAI—a Microsoft Corp. -backed research lab that roughly two months earlier unveiled its text-to-image generator Dall-E2. 

OpenAI CEO Sam Altman, a tech investor and former president of startup accelerator, told The Wall Street Journal in December that the capabilities were still “incredibly embryonic right now,” but are inching developers closer to achieving artificial general intelligence—the ability of software to learn and understand tasks in much the same way that a human can. With deep-pocketed backers like Microsoft, the goal may not be so far fetched. 

Full of Protein

In July, researchers at the DeepMind Technologies artificial-intelligence lab, a subsidiary of Google parent Alphabet Inc., said they had predicted the structure of nearly all known proteins, a significant advance in biology that promises to accelerate drug discovery and help address problems such as sustainability and food insecurity. Using an AI algorithm called AlphaFold, the London-based lab said it had expanded its database of predicted protein structures to 214 million, up from 1 million as of December 2021, representing all proteins known to science, including proteins found in animals, plants, bacteria and other organisms. 

Unlocking Limitless Energy

The U.S. Energy Department in December said that scientists had achieved a breakthrough in research on nuclear fusion, bringing them one step closer to possibly changing the future of clean energy. At a press conference from DOE headquarters in Washington, D.C., Energy Secretary Jennifer Granholm said a controlled fusion reaction at Lawrence Livermore National Laboratory in Livermore, Calif., produced more energy than it consumed. Though researchers say the commercial application of nuclear fusion likely remains years and perhaps decades away, the technology might one day help fight climate change.

Quantum Leaps and Bounds

International Business Machines Corp. announced in November that it has created a more powerful quantum computing chip, the next step in its yearslong effort to build quantum machines capable of delivering business value to companies. The 433-qubit Osprey chip, unveiled at IBM’s annual Quantum Summit in New York, has more than three times as many qubits as the 127-qubit Eagle chip it introduced last year. 

Baidu Inc., a Chinese internet pioneer known for its Google-like search engine, said in late August that it had built its own version of a quantum computer, an experimental device that exploits the quirks of quantum physics to perform calculations at speeds far beyond those of conventional electronic computers.

In March, Sandbox AQ, a software startup developing quantum-computing and artificial-intelligence tools for commercial use, officially spun off from Alphabet’s Google to become a stand-alone company. While full commercial-grade quantum computing is still years away, industry analysts say, Sandbox’s goal is to develop quantum-enabled software tools for use by companies in sectors such as telecommunications, financial services, healthcare and government.

Write to Angus Loten at [email protected]